feat!: enable read cache and write cache when using remote object stores (#5093)

* feat: enable read cache and write cache when using remote object stores

* feat: make read cache be aware of remote store names

* chore: docs

* chore: apply review suggestions

* chore: trim write cache path

---------

Co-authored-by: Yingwen <realevenyag@gmail.com>

.env.example

@@ -14,10 +14,11 @@ GT_AZBLOB_CONTAINER=AZBLOB container
 GT_AZBLOB_ACCOUNT_NAME=AZBLOB account name
 GT_AZBLOB_ACCOUNT_KEY=AZBLOB account key
 GT_AZBLOB_ENDPOINT=AZBLOB endpoint
-# Settings for gcs test 
-GT_GCS_BUCKET = GCS bucket 
+# Settings for gcs test
+GT_GCS_BUCKET = GCS bucket
 GT_GCS_SCOPE  = GCS scope
-GT_GCS_CREDENTIAL_PATH = GCS credential path 
+GT_GCS_CREDENTIAL_PATH = GCS credential path
+GT_GCS_CREDENTIAL = GCS credential
 GT_GCS_ENDPOINT = GCS end point
 # Settings for kafka wal test
 GT_KAFKA_ENDPOINTS = localhost:9092

.github/actions/build-dev-builder-images/action.yml

@@ -50,7 +50,7 @@ runs:
           BUILDX_MULTI_PLATFORM_BUILD=all \
           IMAGE_REGISTRY=${{ inputs.dockerhub-image-registry }} \
           IMAGE_NAMESPACE=${{ inputs.dockerhub-image-namespace }} \
-          IMAGE_TAG=${{ inputs.version }}
+          DEV_BUILDER_IMAGE_TAG=${{ inputs.version }}
 
     - name: Build and push dev-builder-centos image
       shell: bash
@@ -61,7 +61,7 @@ runs:
           BUILDX_MULTI_PLATFORM_BUILD=amd64 \
           IMAGE_REGISTRY=${{ inputs.dockerhub-image-registry }} \
           IMAGE_NAMESPACE=${{ inputs.dockerhub-image-namespace }} \
-          IMAGE_TAG=${{ inputs.version }}
+          DEV_BUILDER_IMAGE_TAG=${{ inputs.version }}
 
     - name: Build and push dev-builder-android image # Only build image for amd64 platform.
       shell: bash
@@ -71,6 +71,6 @@ runs:
           BASE_IMAGE=android \
           IMAGE_REGISTRY=${{ inputs.dockerhub-image-registry }} \
           IMAGE_NAMESPACE=${{ inputs.dockerhub-image-namespace }} \
-          IMAGE_TAG=${{ inputs.version }} && \
+          DEV_BUILDER_IMAGE_TAG=${{ inputs.version }} && \
 
         docker push ${{ inputs.dockerhub-image-registry }}/${{ inputs.dockerhub-image-namespace }}/dev-builder-android:${{ inputs.version }}

.github/actions/build-linux-artifacts/action.yml

@@ -17,6 +17,12 @@ inputs:
     description: Enable dev mode, only build standard greptime
     required: false
     default: "false"
+  image-namespace:
+    description: Image Namespace
+    required: true
+  image-registry:
+    description: Image Registry
+    required: true
   working-dir:
     description: Working directory to build the artifacts
     required: false
@@ -31,8 +37,8 @@ runs:
       run: |
         cd ${{ inputs.working-dir }} && \
         make run-it-in-container BUILD_JOBS=4 \
-        IMAGE_NAMESPACE=i8k6a5e1/greptime \
-        IMAGE_REGISTRY=public.ecr.aws
+        IMAGE_NAMESPACE=${{ inputs.image-namespace }} \
+        IMAGE_REGISTRY=${{ inputs.image-registry }}
 
     - name: Upload sqlness logs
       if: ${{ failure() && inputs.disable-run-tests == 'false' }} # Only upload logs when the integration tests failed.
@@ -51,8 +57,8 @@ runs:
         artifacts-dir: greptime-linux-${{ inputs.arch }}-pyo3-${{ inputs.version }}
         version: ${{ inputs.version }}
         working-dir: ${{ inputs.working-dir }}
-        image-registry: public.ecr.aws
-        image-namespace: i8k6a5e1/greptime        
+        image-registry: ${{ inputs.image-registry }}
+        image-namespace: ${{ inputs.image-namespace }}
 
     - name: Build greptime without pyo3
       if: ${{ inputs.dev-mode == 'false' }}
@@ -64,8 +70,8 @@ runs:
         artifacts-dir: greptime-linux-${{ inputs.arch }}-${{ inputs.version }}
         version: ${{ inputs.version }}
         working-dir: ${{ inputs.working-dir }}
-        image-registry: public.ecr.aws
-        image-namespace: i8k6a5e1/greptime
+        image-registry: ${{ inputs.image-registry }}
+        image-namespace: ${{ inputs.image-namespace }}
 
     - name: Clean up the target directory # Clean up the target directory for the centos7 base image, or it will still use the objects of last build.
       shell: bash
@@ -82,8 +88,8 @@ runs:
         artifacts-dir: greptime-linux-${{ inputs.arch }}-centos-${{ inputs.version }}
         version: ${{ inputs.version }}
         working-dir: ${{ inputs.working-dir }}
-        image-registry: public.ecr.aws
-        image-namespace: i8k6a5e1/greptime
+        image-registry: ${{ inputs.image-registry }}
+        image-namespace: ${{ inputs.image-namespace }}
 
     - name: Build greptime on android base image
       uses: ./.github/actions/build-greptime-binary
@@ -94,5 +100,5 @@ runs:
         version: ${{ inputs.version }}
         working-dir: ${{ inputs.working-dir }}
         build-android-artifacts: true
-        image-registry: public.ecr.aws
-        image-namespace: i8k6a5e1/greptime
+        image-registry: ${{ inputs.image-registry }}
+        image-namespace: ${{ inputs.image-namespace }}

.github/actions/build-macos-artifacts/action.yml

@@ -4,9 +4,6 @@ inputs:
   arch:
     description: Architecture to build
     required: true
-  rust-toolchain:
-    description: Rust toolchain to use
-    required: true
   cargo-profile:
     description: Cargo profile to build
     required: true
@@ -43,10 +40,9 @@ runs:
         brew install protobuf
 
     - name: Install rust toolchain
-      uses: dtolnay/rust-toolchain@master
+      uses: actions-rust-lang/setup-rust-toolchain@v1
       with:
-        toolchain: ${{ inputs.rust-toolchain }}
-        targets: ${{ inputs.arch }}
+        target: ${{ inputs.arch }}
 
     - name: Start etcd # For integration tests.
       if: ${{ inputs.disable-run-tests == 'false' }}

.github/actions/build-windows-artifacts/action.yml

@@ -4,9 +4,6 @@ inputs:
   arch:
     description: Architecture to build
     required: true
-  rust-toolchain:
-    description: Rust toolchain to use
-    required: true
   cargo-profile:
     description: Cargo profile to build
     required: true
@@ -28,10 +25,9 @@ runs:
     - uses: arduino/setup-protoc@v3
 
     - name: Install rust toolchain
-      uses: dtolnay/rust-toolchain@master
+      uses: actions-rust-lang/setup-rust-toolchain@v1
       with:
-        toolchain: ${{ inputs.rust-toolchain }}
-        targets: ${{ inputs.arch }}
+        target: ${{ inputs.arch }}
         components: llvm-tools-preview
 
     - name: Rust Cache
@@ -44,7 +40,7 @@ runs:
 
     - name: Install PyArrow Package
       shell: pwsh
-      run: pip install pyarrow
+      run: pip install pyarrow numpy
 
     - name: Install WSL distribution
       uses: Vampire/setup-wsl@v2

.github/actions/setup-etcd-cluster/action.yml

@@ -18,6 +18,8 @@ runs:
         --set replicaCount=${{ inputs.etcd-replicas }} \
         --set resources.requests.cpu=50m \
         --set resources.requests.memory=128Mi \
+        --set resources.limits.cpu=1500m \
+        --set resources.limits.memory=2Gi \
         --set auth.rbac.create=false \
         --set auth.rbac.token.enabled=false \
         --set persistence.size=2Gi \

.github/actions/setup-greptimedb-cluster/action.yml

@@ -8,7 +8,7 @@ inputs:
     default: 2
     description: "Number of Datanode replicas"
   meta-replicas:
-    default: 3
+    default: 1
     description: "Number of Metasrv replicas"
   image-registry: 
     default: "docker.io"
@@ -58,7 +58,7 @@ runs:
         --set image.tag=${{ inputs.image-tag }} \
         --set base.podTemplate.main.resources.requests.cpu=50m \
         --set base.podTemplate.main.resources.requests.memory=256Mi \
-        --set base.podTemplate.main.resources.limits.cpu=1000m \
+        --set base.podTemplate.main.resources.limits.cpu=2000m \
         --set base.podTemplate.main.resources.limits.memory=2Gi \
         --set frontend.replicas=${{ inputs.frontend-replicas }} \
         --set datanode.replicas=${{ inputs.datanode-replicas }} \

.github/actions/setup-postgres-cluster/action.yml

@@ -0,0 +1,30 @@
+name: Setup PostgreSQL
+description: Deploy PostgreSQL on Kubernetes
+inputs:
+  postgres-replicas:
+    default: 1
+    description: "Number of PostgreSQL replicas"
+  namespace:
+    default: "postgres-namespace"
+  postgres-version:
+    default: "14.2"
+    description: "PostgreSQL version"
+  storage-size:
+    default: "1Gi"
+    description: "Storage size for PostgreSQL"
+
+runs:
+  using: composite
+  steps:
+  - name: Install PostgreSQL
+    shell: bash
+    run: |
+      helm upgrade \
+        --install postgresql oci://registry-1.docker.io/bitnamicharts/postgresql \
+        --set replicaCount=${{ inputs.postgres-replicas }} \
+        --set image.tag=${{ inputs.postgres-version }} \
+        --set persistence.size=${{ inputs.storage-size }} \
+        --set postgresql.username=greptimedb \
+        --set postgresql.password=admin \
+        --create-namespace \
+        -n ${{ inputs.namespace }}

.github/actions/start-runner/action.yml

@@ -38,7 +38,7 @@ runs:
   steps:
     - name: Configure AWS credentials
       if: startsWith(inputs.runner, 'ec2')
-      uses: aws-actions/configure-aws-credentials@v2
+      uses: aws-actions/configure-aws-credentials@v4
       with:
         aws-access-key-id: ${{ inputs.aws-access-key-id }}
         aws-secret-access-key: ${{ inputs.aws-secret-access-key }}

.github/actions/stop-runner/action.yml

@@ -25,7 +25,7 @@ runs:
   steps:
     - name: Configure AWS credentials
       if: ${{ inputs.label && inputs.ec2-instance-id }}
-      uses: aws-actions/configure-aws-credentials@v2
+      uses: aws-actions/configure-aws-credentials@v4
       with:
         aws-access-key-id: ${{ inputs.aws-access-key-id }}
         aws-secret-access-key: ${{ inputs.aws-secret-access-key }}

.github/cargo-blacklist.txt

@@ -0,0 +1,2 @@
+native-tls
+openssl

.github/scripts/check-install-script.sh

@@ -0,0 +1,14 @@
+#!/bin/sh
+
+set -e
+
+# Get the latest version of github.com/GreptimeTeam/greptimedb
+VERSION=$(curl -s https://api.github.com/repos/GreptimeTeam/greptimedb/releases/latest | jq -r '.tag_name')
+
+echo "Downloading the latest version: $VERSION"
+
+# Download the install script
+curl -fsSL https://raw.githubusercontent.com/greptimeteam/greptimedb/main/scripts/install.sh | sh -s $VERSION
+
+# Execute the `greptime` command
+./greptime --version

.github/workflows/apidoc.yml

@@ -12,9 +12,6 @@ on:
 
 name: Build API docs
 
-env:
-  RUST_TOOLCHAIN: nightly-2024-04-20
-
 jobs:
   apidoc:
     runs-on: ubuntu-20.04
@@ -23,9 +20,7 @@ jobs:
     - uses: arduino/setup-protoc@v3
       with:
         repo-token: ${{ secrets.GITHUB_TOKEN }}
-    - uses: dtolnay/rust-toolchain@master
-      with:
-        toolchain: ${{ env.RUST_TOOLCHAIN }}
+    - uses: actions-rust-lang/setup-rust-toolchain@v1
     - run: cargo doc --workspace --no-deps --document-private-items
     - run: |
         cat <<EOF > target/doc/index.html

.github/workflows/dependency-check.yml

@@ -0,0 +1,36 @@
+name: Check Dependencies
+
+on:
+  push:
+    branches:
+      - main
+  pull_request:
+    branches:
+      - main
+
+jobs:
+  check-dependencies:
+    runs-on: ubuntu-latest
+
+    steps:
+    - name: Checkout code
+      uses: actions/checkout@v4
+
+    - name: Set up Rust
+      uses: actions-rust-lang/setup-rust-toolchain@v1
+
+    - name: Run cargo tree
+      run: cargo tree --prefix none > dependencies.txt
+
+    - name: Extract dependency names
+      run: awk '{print $1}' dependencies.txt > dependency_names.txt
+
+    - name: Check for blacklisted crates
+      run: |
+        while read -r dep; do
+          if grep -qFx "$dep" dependency_names.txt; then
+            echo "Blacklisted crate '$dep' found in dependencies."
+            exit 1
+          fi
+        done < .github/cargo-blacklist.txt
+        echo "No blacklisted crates found."

.github/workflows/dev-build.yml

@@ -177,6 +177,8 @@ jobs:
           disable-run-tests: ${{ env.DISABLE_RUN_TESTS }}
           dev-mode: true # Only build the standard greptime binary.
           working-dir: ${{ env.CHECKOUT_GREPTIMEDB_PATH }}
+          image-registry: ${{ vars.ECR_IMAGE_REGISTRY }}
+          image-namespace: ${{ vars.ECR_IMAGE_NAMESPACE }}
 
   build-linux-arm64-artifacts:
     name: Build linux-arm64 artifacts
@@ -206,6 +208,8 @@ jobs:
           disable-run-tests: ${{ env.DISABLE_RUN_TESTS }}
           dev-mode: true # Only build the standard greptime binary.
           working-dir: ${{ env.CHECKOUT_GREPTIMEDB_PATH }}
+          image-registry: ${{ vars.ECR_IMAGE_REGISTRY }}
+          image-namespace: ${{ vars.ECR_IMAGE_NAMESPACE }}
 
   release-images-to-dockerhub:
     name: Build and push images to DockerHub

.github/workflows/develop.yml

@@ -29,9 +29,6 @@ concurrency:
   group: ${{ github.workflow }}-${{ github.head_ref || github.run_id }}
   cancel-in-progress: true
 
-env:
-  RUST_TOOLCHAIN: nightly-2024-04-20
-
 jobs:
   check-typos-and-docs:
     name: Check typos and docs
@@ -64,9 +61,7 @@ jobs:
       - uses: arduino/setup-protoc@v3
         with:
           repo-token: ${{ secrets.GITHUB_TOKEN }}
-      - uses: dtolnay/rust-toolchain@master
-        with:
-          toolchain: ${{ env.RUST_TOOLCHAIN }}
+      - uses: actions-rust-lang/setup-rust-toolchain@v1
       - name: Rust Cache
         uses: Swatinem/rust-cache@v2
         with:
@@ -82,9 +77,7 @@ jobs:
     timeout-minutes: 60
     steps:
       - uses: actions/checkout@v4
-      - uses: dtolnay/rust-toolchain@master
-        with:
-          toolchain: stable
+      - uses: actions-rust-lang/setup-rust-toolchain@v1
       - name: Rust Cache
         uses: Swatinem/rust-cache@v2
         with:
@@ -107,9 +100,7 @@ jobs:
       - uses: arduino/setup-protoc@v3
         with:
           repo-token: ${{ secrets.GITHUB_TOKEN }}
-      - uses: dtolnay/rust-toolchain@master
-        with:
-          toolchain: ${{ env.RUST_TOOLCHAIN }}
+      - uses: actions-rust-lang/setup-rust-toolchain@v1
       - uses: Swatinem/rust-cache@v2
         with:
           # Shares across multiple jobs
@@ -161,9 +152,7 @@ jobs:
       - uses: arduino/setup-protoc@v3
         with:
           repo-token: ${{ secrets.GITHUB_TOKEN }}
-      - uses: dtolnay/rust-toolchain@master
-        with:
-          toolchain: ${{ env.RUST_TOOLCHAIN }}
+      - uses: actions-rust-lang/setup-rust-toolchain@v1
       - name: Rust Cache
         uses: Swatinem/rust-cache@v2
         with:
@@ -181,7 +170,7 @@ jobs:
           name: bins
           path: .
       - name: Unzip binaries
-        run: | 
+        run: |
           tar -xvf ./bins.tar.gz
           rm ./bins.tar.gz
       - name: Run GreptimeDB
@@ -221,9 +210,7 @@ jobs:
       - uses: arduino/setup-protoc@v3
         with:
           repo-token: ${{ secrets.GITHUB_TOKEN }}
-      - uses: dtolnay/rust-toolchain@master
-        with:
-          toolchain: ${{ env.RUST_TOOLCHAIN }}
+      - uses: actions-rust-lang/setup-rust-toolchain@v1
       - name: Rust Cache
         uses: Swatinem/rust-cache@v2
         with:
@@ -274,9 +261,7 @@ jobs:
       - uses: arduino/setup-protoc@v3
         with:
           repo-token: ${{ secrets.GITHUB_TOKEN }}
-      - uses: dtolnay/rust-toolchain@master
-        with:
-          toolchain: ${{ env.RUST_TOOLCHAIN }}
+      - uses: actions-rust-lang/setup-rust-toolchain@v1
       - uses: Swatinem/rust-cache@v2
         with:
           # Shares across multiple jobs
@@ -284,10 +269,17 @@ jobs:
       - name: Install cargo-gc-bin
         shell: bash
         run: cargo install cargo-gc-bin
+      - name: Check aws-lc-sys will not build
+        shell: bash
+        run: |
+             if cargo tree -i aws-lc-sys -e features | grep -q aws-lc-sys; then
+               echo "Found aws-lc-sys, which has compilation problems on older gcc versions. Please replace it with ring until its building experience improves."
+               exit 1
+             fi
       - name: Build greptime bianry
         shell: bash
         # `cargo gc` will invoke `cargo build` with specified args
-        run: cargo gc --profile ci -- --bin greptime 
+        run: cargo gc --profile ci -- --bin greptime
       - name: Pack greptime binary
         shell: bash
         run: |
@@ -301,7 +293,7 @@ jobs:
           artifacts-dir: bin
           version: current
 
-  distributed-fuzztest: 
+  distributed-fuzztest:
     name: Fuzz Test (Distributed, ${{ matrix.mode.name }}, ${{ matrix.target }})
     runs-on: ubuntu-latest
     needs:  build-greptime-ci
@@ -338,13 +330,13 @@ jobs:
         uses: ./.github/actions/setup-kafka-cluster
       - name: Setup Etcd cluser
         uses: ./.github/actions/setup-etcd-cluster
+      - name: Setup Postgres cluser
+        uses: ./.github/actions/setup-postgres-cluster
       # Prepares for fuzz tests
       - uses: arduino/setup-protoc@v3
         with:
           repo-token: ${{ secrets.GITHUB_TOKEN }}
-      - uses: dtolnay/rust-toolchain@master
-        with:
-          toolchain: ${{ env.RUST_TOOLCHAIN }}
+      - uses: actions-rust-lang/setup-rust-toolchain@v1
       - name: Rust Cache
         uses: Swatinem/rust-cache@v2
         with:
@@ -414,12 +406,12 @@ jobs:
       - name: Describe Nodes
         if: failure()
         shell: bash
-        run: | 
-          kubectl describe nodes      
+        run: |
+          kubectl describe nodes
       - name: Export kind logs
         if: failure()
         shell: bash
-        run: | 
+        run: |
           kind export logs /tmp/kind
       - name: Upload logs
         if: failure()
@@ -431,25 +423,38 @@ jobs:
       - name: Delete cluster
         if: success()
         shell: bash
-        run: | 
+        run: |
           kind delete cluster
           docker stop $(docker ps -a -q)
           docker rm $(docker ps -a -q)
           docker system prune -f
 
-  distributed-fuzztest-with-chaos: 
+  distributed-fuzztest-with-chaos:
     name: Fuzz Test with Chaos (Distributed, ${{ matrix.mode.name }}, ${{ matrix.target }})
     runs-on: ubuntu-latest
     needs:  build-greptime-ci
     timeout-minutes: 60
     strategy:
       matrix:
-        target: ["fuzz_migrate_mito_regions", "fuzz_failover_mito_regions", "fuzz_failover_metric_regions"]
-        mode: 
+        target: ["fuzz_migrate_mito_regions", "fuzz_migrate_metric_regions", "fuzz_failover_mito_regions", "fuzz_failover_metric_regions"]
+        mode:
           - name: "Remote WAL"
             minio: true
             kafka: true
             values: "with-remote-wal.yaml"
+        include:
+          - target: "fuzz_migrate_mito_regions"
+            mode:
+              name: "Local WAL"
+              minio: true
+              kafka: false
+              values: "with-minio.yaml"
+          - target: "fuzz_migrate_metric_regions"
+            mode:
+              name: "Local WAL"
+              minio: true
+              kafka: false
+              values: "with-minio.yaml"
     steps:
       - name: Remove unused software
         run: |
@@ -476,13 +481,13 @@ jobs:
         uses: ./.github/actions/setup-kafka-cluster
       - name: Setup Etcd cluser
         uses: ./.github/actions/setup-etcd-cluster
+      - name: Setup Postgres cluser
+        uses: ./.github/actions/setup-postgres-cluster
       # Prepares for fuzz tests
       - uses: arduino/setup-protoc@v3
         with:
           repo-token: ${{ secrets.GITHUB_TOKEN }}
-      - uses: dtolnay/rust-toolchain@master
-        with:
-          toolchain: ${{ env.RUST_TOOLCHAIN }}
+      - uses: actions-rust-lang/setup-rust-toolchain@v1
       - name: Rust Cache
         uses: Swatinem/rust-cache@v2
         with:
@@ -538,7 +543,7 @@ jobs:
         with:
           image-registry: localhost:5001
           values-filename: ${{ matrix.mode.values }}
-          enable-region-failover: true
+          enable-region-failover: ${{ matrix.mode.kafka }}
       - name: Port forward (mysql)
         run: |
           kubectl port-forward service/my-greptimedb-frontend 4002:4002 -n my-greptimedb&
@@ -553,12 +558,12 @@ jobs:
       - name: Describe Nodes
         if: failure()
         shell: bash
-        run: | 
-          kubectl describe nodes      
+        run: |
+          kubectl describe nodes
       - name: Export kind logs
         if: failure()
         shell: bash
-        run: | 
+        run: |
           kind export logs /tmp/kind
       - name: Upload logs
         if: failure()
@@ -570,7 +575,7 @@ jobs:
       - name: Delete cluster
         if: success()
         shell: bash
-        run: | 
+        run: |
           kind delete cluster
           docker stop $(docker ps -a -q)
           docker rm $(docker ps -a -q)
@@ -593,6 +598,10 @@ jobs:
     timeout-minutes: 60
     steps:
       - uses: actions/checkout@v4
+      - if: matrix.mode.kafka
+        name: Setup kafka server
+        working-directory: tests-integration/fixtures/kafka
+        run: docker compose -f docker-compose-standalone.yml up -d --wait
       - name: Download pre-built binaries
         uses: actions/download-artifact@v4
         with:
@@ -600,10 +609,6 @@ jobs:
           path: .
       - name: Unzip binaries
         run: tar -xvf ./bins.tar.gz
-      - if: matrix.mode.kafka
-        name: Setup kafka server
-        working-directory: tests-integration/fixtures/kafka
-        run: docker compose -f docker-compose-standalone.yml up -d --wait
       - name: Run sqlness
         run: RUST_BACKTRACE=1 ./bins/sqlness-runner ${{ matrix.mode.opts }} -c ./tests/cases --bins-dir ./bins --preserve-state
       - name: Upload sqlness logs
@@ -623,17 +628,16 @@ jobs:
       - uses: arduino/setup-protoc@v3
         with:
           repo-token: ${{ secrets.GITHUB_TOKEN }}
-      - uses: dtolnay/rust-toolchain@master
+      - uses: actions-rust-lang/setup-rust-toolchain@v1
         with:
-          toolchain: ${{ env.RUST_TOOLCHAIN }}
           components: rustfmt
       - name: Rust Cache
         uses: Swatinem/rust-cache@v2
         with:
           # Shares across multiple jobs
           shared-key: "check-rust-fmt"
-      - name: Run cargo fmt
-        run: cargo fmt --all -- --check
+      - name: Check format
+        run: make fmt-check
 
   clippy:
     name: Clippy
@@ -644,9 +648,8 @@ jobs:
       - uses: arduino/setup-protoc@v3
         with:
           repo-token: ${{ secrets.GITHUB_TOKEN }}
-      - uses: dtolnay/rust-toolchain@master
+      - uses: actions-rust-lang/setup-rust-toolchain@v1
         with:
-          toolchain: ${{ env.RUST_TOOLCHAIN }}
           components: clippy
       - name: Rust Cache
         uses: Swatinem/rust-cache@v2
@@ -670,9 +673,8 @@ jobs:
         with:
           version: "14.0"
       - name: Install toolchain
-        uses: dtolnay/rust-toolchain@master
+        uses: actions-rust-lang/setup-rust-toolchain@v1
         with:
-          toolchain: ${{ env.RUST_TOOLCHAIN }}
           components: llvm-tools-preview
       - name: Rust Cache
         uses: Swatinem/rust-cache@v2
@@ -692,7 +694,7 @@ jobs:
         with:
           python-version: '3.10'
       - name: Install PyArrow Package
-        run: pip install pyarrow
+        run: pip install pyarrow numpy
       - name: Setup etcd server
         working-directory: tests-integration/fixtures/etcd
         run: docker compose -f docker-compose-standalone.yml up -d --wait
@@ -702,6 +704,9 @@ jobs:
       - name: Setup minio
         working-directory: tests-integration/fixtures/minio
         run: docker compose -f docker-compose-standalone.yml up -d --wait
+      - name: Setup postgres server
+        working-directory: tests-integration/fixtures/postgres
+        run: docker compose -f docker-compose-standalone.yml up -d --wait
       - name: Run nextest cases
         run: cargo llvm-cov nextest --workspace --lcov --output-path lcov.info -F pyo3_backend -F dashboard
         env:
@@ -718,7 +723,9 @@ jobs:
           GT_MINIO_REGION: us-west-2
           GT_MINIO_ENDPOINT_URL: http://127.0.0.1:9000
           GT_ETCD_ENDPOINTS: http://127.0.0.1:2379
+          GT_POSTGRES_ENDPOINTS: postgres://greptimedb:admin@127.0.0.1:5432/postgres
           GT_KAFKA_ENDPOINTS: 127.0.0.1:9092
+          GT_KAFKA_SASL_ENDPOINTS: 127.0.0.1:9093
           UNITTEST_LOG_DIR: "__unittest_logs"
       - name: Codecov upload
         uses: codecov/codecov-action@v4

.github/workflows/nightly-build.yml

@@ -154,6 +154,8 @@ jobs:
           cargo-profile: ${{ env.CARGO_PROFILE }}
           version: ${{ needs.allocate-runners.outputs.version }}
           disable-run-tests: ${{ env.DISABLE_RUN_TESTS }}
+          image-registry: ${{ vars.ECR_IMAGE_REGISTRY }}
+          image-namespace: ${{ vars.ECR_IMAGE_NAMESPACE }}
 
   build-linux-arm64-artifacts:
     name: Build linux-arm64 artifacts
@@ -173,6 +175,8 @@ jobs:
           cargo-profile: ${{ env.CARGO_PROFILE }}
           version: ${{ needs.allocate-runners.outputs.version }}
           disable-run-tests: ${{ env.DISABLE_RUN_TESTS }}
+          image-registry: ${{ vars.ECR_IMAGE_REGISTRY }}
+          image-namespace: ${{ vars.ECR_IMAGE_NAMESPACE }}
 
   release-images-to-dockerhub:
     name: Build and push images to DockerHub

.github/workflows/nightly-ci.yml

@@ -9,9 +9,6 @@ concurrency:
   group: ${{ github.workflow }}-${{ github.head_ref || github.run_id }}
   cancel-in-progress: true
 
-env:
-  RUST_TOOLCHAIN: nightly-2024-04-20
-
 permissions:
   issues: write
 
@@ -25,6 +22,10 @@ jobs:
         uses: actions/checkout@v4
         with:
           fetch-depth: 0
+
+      - name: Check install.sh
+        run: ./.github/scripts/check-install-script.sh
+
       - name: Run sqlness test
         uses: ./.github/actions/sqlness-test
         with:
@@ -33,6 +34,13 @@ jobs:
           aws-region: ${{ vars.AWS_CI_TEST_BUCKET_REGION }}
           aws-access-key-id: ${{ secrets.AWS_CI_TEST_ACCESS_KEY_ID }}
           aws-secret-access-key: ${{ secrets.AWS_CI_TEST_SECRET_ACCESS_KEY }}
+      - name: Upload sqlness logs
+        if: failure()
+        uses: actions/upload-artifact@v4
+        with:
+          name: sqlness-logs-kind
+          path: /tmp/kind/
+          retention-days: 3
 
   sqlness-windows:
     name: Sqlness tests on Windows
@@ -45,9 +53,7 @@ jobs:
       - uses: arduino/setup-protoc@v3
         with:
           repo-token: ${{ secrets.GITHUB_TOKEN }}
-      - uses: dtolnay/rust-toolchain@master
-        with:
-          toolchain: ${{ env.RUST_TOOLCHAIN }}
+      - uses: actions-rust-lang/setup-rust-toolchain@v1
       - name: Rust Cache
         uses: Swatinem/rust-cache@v2
       - name: Run sqlness
@@ -55,7 +61,7 @@ jobs:
         env:
           SQLNESS_OPTS: "--preserve-state"
       - name: Upload sqlness logs
-        if: always()
+        if: failure()
         uses: actions/upload-artifact@v4
         with:
           name: sqlness-logs
@@ -78,9 +84,8 @@ jobs:
         with:
           version: "14.0"
       - name: Install Rust toolchain
-        uses: dtolnay/rust-toolchain@master
+        uses: actions-rust-lang/setup-rust-toolchain@v1
         with:
-          toolchain: ${{ env.RUST_TOOLCHAIN }}
           components: llvm-tools-preview
       - name: Rust Cache
         uses: Swatinem/rust-cache@v2
@@ -91,7 +96,7 @@ jobs:
         with:
           python-version: "3.10"
       - name: Install PyArrow Package
-        run: pip install pyarrow
+        run: pip install pyarrow numpy
       - name: Install WSL distribution
         uses: Vampire/setup-wsl@v2
         with:

.github/workflows/release-dev-builder-images.yaml

@@ -1,12 +1,14 @@
 name: Release dev-builder images
 
 on:
+  push:
+    branches:
+      - main
+    paths:
+      - rust-toolchain.toml
+      - 'docker/dev-builder/**'
   workflow_dispatch: # Allows you to run this workflow manually.
     inputs:
-      version:
-        description: Version of the dev-builder
-        required: false
-        default: latest
       release_dev_builder_ubuntu_image:
         type: boolean
         description: Release dev-builder-ubuntu image
@@ -28,22 +30,103 @@ jobs:
     name: Release dev builder images
     if: ${{ inputs.release_dev_builder_ubuntu_image || inputs.release_dev_builder_centos_image || inputs.release_dev_builder_android_image }} # Only manually trigger this job.
     runs-on: ubuntu-20.04-16-cores
+    outputs:
+      version: ${{ steps.set-version.outputs.version }}
     steps:
       - name: Checkout
         uses: actions/checkout@v4
         with:
           fetch-depth: 0
 
+      - name: Configure build image version
+        id: set-version
+        shell: bash
+        run: |
+          commitShortSHA=`echo ${{ github.sha }} | cut -c1-8`
+          buildTime=`date +%Y%m%d%H%M%S`
+          BUILD_VERSION="$commitShortSHA-$buildTime"
+          RUST_TOOLCHAIN_VERSION=$(cat rust-toolchain.toml | grep -Eo '[0-9]{4}-[0-9]{2}-[0-9]{2}')
+          IMAGE_VERSION="${RUST_TOOLCHAIN_VERSION}-${BUILD_VERSION}"
+          echo "VERSION=${IMAGE_VERSION}" >> $GITHUB_ENV
+          echo "version=$IMAGE_VERSION" >> $GITHUB_OUTPUT
+
       - name: Build and push dev builder images
         uses: ./.github/actions/build-dev-builder-images
         with:
-          version: ${{ inputs.version }}
+          version: ${{ env.VERSION }}
           dockerhub-image-registry-username: ${{ secrets.DOCKERHUB_USERNAME }}
           dockerhub-image-registry-token: ${{ secrets.DOCKERHUB_TOKEN }}
           build-dev-builder-ubuntu: ${{ inputs.release_dev_builder_ubuntu_image }}
           build-dev-builder-centos: ${{ inputs.release_dev_builder_centos_image }}
           build-dev-builder-android: ${{ inputs.release_dev_builder_android_image }}
 
+  release-dev-builder-images-ecr:
+    name: Release dev builder images to AWS ECR
+    runs-on: ubuntu-20.04
+    needs: [
+      release-dev-builder-images
+    ]
+    steps:
+      - name: Configure AWS credentials
+        uses: aws-actions/configure-aws-credentials@v4
+        with:
+          aws-access-key-id: ${{ secrets.AWS_ECR_ACCESS_KEY_ID }}
+          aws-secret-access-key: ${{ secrets.AWS_ECR_SECRET_ACCESS_KEY }}
+          aws-region: ${{ vars.ECR_REGION }}
+
+      - name: Login to Amazon ECR
+        id: login-ecr-public
+        uses: aws-actions/amazon-ecr-login@v2
+        env:
+          AWS_REGION: ${{ vars.ECR_REGION }}
+        with:
+          registry-type: public
+
+      - name: Push dev-builder-ubuntu image
+        shell: bash
+        if: ${{ inputs.release_dev_builder_ubuntu_image }}
+        run: |
+          docker run -v "${DOCKER_CONFIG:-$HOME/.docker}:/root/.docker:ro" \
+            -e "REGISTRY_AUTH_FILE=/root/.docker/config.json" \
+            quay.io/skopeo/stable:latest \
+            copy -a docker://docker.io/${{ vars.IMAGE_NAMESPACE }}/dev-builder-ubuntu:${{ needs.release-dev-builder-images.outputs.version }} \
+            docker://${{ vars.ECR_IMAGE_REGISTRY }}/${{ vars.ECR_IMAGE_NAMESPACE }}/dev-builder-ubuntu:${{ needs.release-dev-builder-images.outputs.version }}
+
+          docker run -v "${DOCKER_CONFIG:-$HOME/.docker}:/root/.docker:ro" \
+            -e "REGISTRY_AUTH_FILE=/root/.docker/config.json" \
+            quay.io/skopeo/stable:latest \
+            copy -a docker://docker.io/${{ vars.IMAGE_NAMESPACE }}/dev-builder-ubuntu:latest \
+            docker://${{ vars.ECR_IMAGE_REGISTRY }}/${{ vars.ECR_IMAGE_NAMESPACE }}/dev-builder-ubuntu:latest
+      - name: Push dev-builder-centos image
+        shell: bash
+        if: ${{ inputs.release_dev_builder_centos_image }}
+        run: |
+          docker run -v "${DOCKER_CONFIG:-$HOME/.docker}:/root/.docker:ro" \
+            -e "REGISTRY_AUTH_FILE=/root/.docker/config.json" \
+            quay.io/skopeo/stable:latest \
+            copy -a docker://docker.io/${{ vars.IMAGE_NAMESPACE }}/dev-builder-centos:${{ needs.release-dev-builder-images.outputs.version }} \
+            docker://${{ vars.ECR_IMAGE_REGISTRY }}/${{ vars.ECR_IMAGE_NAMESPACE }}/dev-builder-centos:${{ needs.release-dev-builder-images.outputs.version }}
+
+          docker run -v "${DOCKER_CONFIG:-$HOME/.docker}:/root/.docker:ro" \
+            -e "REGISTRY_AUTH_FILE=/root/.docker/config.json" \
+            quay.io/skopeo/stable:latest \
+            copy -a docker://docker.io/${{ vars.IMAGE_NAMESPACE }}/dev-builder-centos:latest \
+            docker://${{ vars.ECR_IMAGE_REGISTRY }}/${{ vars.ECR_IMAGE_NAMESPACE }}/dev-builder-centos:latest
+      - name: Push dev-builder-android image
+        shell: bash
+        if: ${{ inputs.release_dev_builder_android_image }}
+        run: |
+          docker run -v "${DOCKER_CONFIG:-$HOME/.docker}:/root/.docker:ro" \
+            -e "REGISTRY_AUTH_FILE=/root/.docker/config.json" \
+            quay.io/skopeo/stable:latest \
+            copy -a docker://docker.io/${{ vars.IMAGE_NAMESPACE }}/dev-builder-android:${{ needs.release-dev-builder-images.outputs.version }} \
+            docker://${{ vars.ECR_IMAGE_REGISTRY }}/${{ vars.ECR_IMAGE_NAMESPACE }}/dev-builder-android:${{ needs.release-dev-builder-images.outputs.version }}
+
+          docker run -v "${DOCKER_CONFIG:-$HOME/.docker}:/root/.docker:ro" \
+            -e "REGISTRY_AUTH_FILE=/root/.docker/config.json" \
+            quay.io/skopeo/stable:latest \
+            copy -a docker://docker.io/${{ vars.IMAGE_NAMESPACE }}/dev-builder-android:latest \
+            docker://${{ vars.ECR_IMAGE_REGISTRY }}/${{ vars.ECR_IMAGE_NAMESPACE }}/dev-builder-android:latest
   release-dev-builder-images-cn: # Note: Be careful issue: https://github.com/containers/skopeo/issues/1874 and we decide to use the latest stable skopeo container.
     name: Release dev builder images to CN region
     runs-on: ubuntu-20.04
@@ -51,35 +134,39 @@ jobs:
       release-dev-builder-images
     ]
     steps:
+      - name: Login to AliCloud Container Registry
+        uses: docker/login-action@v3
+        with:
+          registry: ${{ vars.ACR_IMAGE_REGISTRY }}
+          username: ${{ secrets.ALICLOUD_USERNAME }}
+          password: ${{ secrets.ALICLOUD_PASSWORD }}
+
       - name: Push dev-builder-ubuntu image
         shell: bash
         if: ${{ inputs.release_dev_builder_ubuntu_image }}
-        env:
-          DST_REGISTRY_USERNAME: ${{ secrets.ALICLOUD_USERNAME }}
-          DST_REGISTRY_PASSWORD: ${{ secrets.ALICLOUD_PASSWORD }}
         run: |
-          docker run quay.io/skopeo/stable:latest copy -a docker://docker.io/${{ vars.IMAGE_NAMESPACE }}/dev-builder-ubuntu:${{ inputs.version }} \
-            --dest-creds "$DST_REGISTRY_USERNAME":"$DST_REGISTRY_PASSWORD" \
-            docker://${{ vars.ACR_IMAGE_REGISTRY }}/${{ vars.IMAGE_NAMESPACE }}/dev-builder-ubuntu:${{ inputs.version }}
+          docker run -v "${DOCKER_CONFIG:-$HOME/.docker}:/root/.docker:ro" \
+            -e "REGISTRY_AUTH_FILE=/root/.docker/config.json" \
+            quay.io/skopeo/stable:latest \
+            copy -a docker://docker.io/${{ vars.IMAGE_NAMESPACE }}/dev-builder-ubuntu:${{ needs.release-dev-builder-images.outputs.version }} \
+            docker://${{ vars.ACR_IMAGE_REGISTRY }}/${{ vars.IMAGE_NAMESPACE }}/dev-builder-ubuntu:${{ needs.release-dev-builder-images.outputs.version }}
 
       - name: Push dev-builder-centos image
         shell: bash
         if: ${{ inputs.release_dev_builder_centos_image }}
-        env:
-          DST_REGISTRY_USERNAME: ${{ secrets.ALICLOUD_USERNAME }}
-          DST_REGISTRY_PASSWORD: ${{ secrets.ALICLOUD_PASSWORD }}
         run: |
-          docker run quay.io/skopeo/stable:latest copy -a docker://docker.io/${{ vars.IMAGE_NAMESPACE }}/dev-builder-centos:${{ inputs.version }} \
-            --dest-creds "$DST_REGISTRY_USERNAME":"$DST_REGISTRY_PASSWORD" \
-            docker://${{ vars.ACR_IMAGE_REGISTRY }}/${{ vars.IMAGE_NAMESPACE }}/dev-builder-centos:${{ inputs.version }}
+          docker run -v "${DOCKER_CONFIG:-$HOME/.docker}:/root/.docker:ro" \
+            -e "REGISTRY_AUTH_FILE=/root/.docker/config.json" \
+            quay.io/skopeo/stable:latest \
+            copy -a docker://docker.io/${{ vars.IMAGE_NAMESPACE }}/dev-builder-centos:${{ needs.release-dev-builder-images.outputs.version }} \
+            docker://${{ vars.ACR_IMAGE_REGISTRY }}/${{ vars.IMAGE_NAMESPACE }}/dev-builder-centos:${{ needs.release-dev-builder-images.outputs.version }}
 
       - name: Push dev-builder-android image
         shell: bash
         if: ${{ inputs.release_dev_builder_android_image }}
-        env:
-          DST_REGISTRY_USERNAME: ${{ secrets.ALICLOUD_USERNAME }}
-          DST_REGISTRY_PASSWORD: ${{ secrets.ALICLOUD_PASSWORD }}
         run: |
-          docker run quay.io/skopeo/stable:latest copy -a docker://docker.io/${{ vars.IMAGE_NAMESPACE }}/dev-builder-android:${{ inputs.version }} \
-            --dest-creds "$DST_REGISTRY_USERNAME":"$DST_REGISTRY_PASSWORD" \
-            docker://${{ vars.ACR_IMAGE_REGISTRY }}/${{ vars.IMAGE_NAMESPACE }}/dev-builder-android:${{ inputs.version }}
+          docker run -v "${DOCKER_CONFIG:-$HOME/.docker}:/root/.docker:ro" \
+            -e "REGISTRY_AUTH_FILE=/root/.docker/config.json" \
+            quay.io/skopeo/stable:latest \
+            copy -a docker://docker.io/${{ vars.IMAGE_NAMESPACE }}/dev-builder-android:${{ needs.release-dev-builder-images.outputs.version }} \
+            docker://${{ vars.ACR_IMAGE_REGISTRY }}/${{ vars.IMAGE_NAMESPACE }}/dev-builder-android:${{ needs.release-dev-builder-images.outputs.version }}

.github/workflows/release.yml

@@ -33,6 +33,7 @@ on:
         description: The runner uses to build linux-arm64 artifacts
         default: ec2-c6g.4xlarge-arm64
         options:
+          - ubuntu-2204-32-cores-arm
           - ec2-c6g.xlarge-arm64 # 4C8G
           - ec2-c6g.2xlarge-arm64 # 8C16G
           - ec2-c6g.4xlarge-arm64 # 16C32G
@@ -82,7 +83,6 @@ on:
 # Use env variables to control all the release process.
 env:
   # The arguments of building greptime.
-  RUST_TOOLCHAIN: nightly-2024-04-20
   CARGO_PROFILE: nightly
 
   # Controls whether to run tests, include unit-test, integration-test and sqlness.
@@ -91,7 +91,7 @@ env:
   # The scheduled version is '${{ env.NEXT_RELEASE_VERSION }}-nightly-YYYYMMDD', like v0.2.0-nigthly-20230313;
   NIGHTLY_RELEASE_PREFIX: nightly
   # Note: The NEXT_RELEASE_VERSION should be modified manually by every formal release.
-  NEXT_RELEASE_VERSION: v0.10.0
+  NEXT_RELEASE_VERSION: v0.11.0
 
 # Permission reference: https://docs.github.com/en/actions/using-jobs/assigning-permissions-to-jobs
 permissions:
@@ -123,6 +123,11 @@ jobs:
         with:
           fetch-depth: 0
 
+      - name: Check Rust toolchain version
+        shell: bash
+        run: |
+          ./scripts/check-builder-rust-version.sh
+
       # The create-version will create a global variable named 'version' in the global workflows.
       # - If it's a tag push release, the version is the tag name(${{ github.ref_name }});
       # - If it's a scheduled release, the version is '${{ env.NEXT_RELEASE_VERSION }}-nightly-$buildTime', like v0.2.0-nigthly-20230313;
@@ -183,6 +188,8 @@ jobs:
           cargo-profile: ${{ env.CARGO_PROFILE }}
           version: ${{ needs.allocate-runners.outputs.version }}
           disable-run-tests: ${{ env.DISABLE_RUN_TESTS }}
+          image-registry: ${{ vars.ECR_IMAGE_REGISTRY }}
+          image-namespace: ${{ vars.ECR_IMAGE_NAMESPACE }}
 
   build-linux-arm64-artifacts:
     name: Build linux-arm64 artifacts
@@ -202,6 +209,8 @@ jobs:
           cargo-profile: ${{ env.CARGO_PROFILE }}
           version: ${{ needs.allocate-runners.outputs.version }}
           disable-run-tests: ${{ env.DISABLE_RUN_TESTS }}
+          image-registry: ${{ vars.ECR_IMAGE_REGISTRY }}
+          image-namespace: ${{ vars.ECR_IMAGE_NAMESPACE }}
 
   build-macos-artifacts:
     name: Build macOS artifacts
@@ -240,11 +249,11 @@ jobs:
       - uses: ./.github/actions/build-macos-artifacts
         with:
           arch: ${{ matrix.arch }}
-          rust-toolchain: ${{ env.RUST_TOOLCHAIN }}
           cargo-profile: ${{ env.CARGO_PROFILE }}
           features: ${{ matrix.features }}
           version: ${{ needs.allocate-runners.outputs.version }}
-          disable-run-tests: ${{ env.DISABLE_RUN_TESTS }}
+          # We decide to disable the integration tests on macOS because it's unnecessary and time-consuming.
+          disable-run-tests: true
           artifacts-dir: ${{ matrix.artifacts-dir-prefix }}-${{ needs.allocate-runners.outputs.version }}
 
       - name: Set build macos result
@@ -283,7 +292,6 @@ jobs:
       - uses: ./.github/actions/build-windows-artifacts
         with:
           arch: ${{ matrix.arch }}
-          rust-toolchain: ${{ env.RUST_TOOLCHAIN }}
           cargo-profile: ${{ env.CARGO_PROFILE }}
           features: ${{ matrix.features }}
           version: ${{ needs.allocate-runners.outputs.version }}

.pre-commit-config.yaml

@@ -17,6 +17,6 @@ repos:
     -    id: fmt
     -    id: clippy
          args: ["--workspace", "--all-targets", "--all-features", "--", "-D", "warnings"]
-         stages: [push]
+         stages: [pre-push]
     -    id: cargo-check
          args: ["--workspace", "--all-targets", "--all-features"]

AUTHOR.md

@@ -7,6 +7,8 @@
 * [NiwakaDev](https://github.com/NiwakaDev)
 * [etolbakov](https://github.com/etolbakov)
 * [irenjj](https://github.com/irenjj)
+* [tisonkun](https://github.com/tisonkun)
+* [Lanqing Yang](https://github.com/lyang24)
 
 ## Team Members (in alphabetical order)
 
@@ -30,7 +32,6 @@
 * [shuiyisong](https://github.com/shuiyisong)
 * [sunchanglong](https://github.com/sunchanglong)
 * [sunng87](https://github.com/sunng87)
-* [tisonkun](https://github.com/tisonkun)
 * [v0y4g3r](https://github.com/v0y4g3r)
 * [waynexia](https://github.com/waynexia)
 * [xtang](https://github.com/xtang)

Cargo.toml

@@ -2,24 +2,28 @@
 members = [
     "src/api",
     "src/auth",
-    "src/catalog",
     "src/cache",
+    "src/catalog",
+    "src/cli",
     "src/client",
     "src/cmd",
     "src/common/base",
     "src/common/catalog",
     "src/common/config",
     "src/common/datasource",
+    "src/common/decimal",
     "src/common/error",
     "src/common/frontend",
     "src/common/function",
-    "src/common/macro",
     "src/common/greptimedb-telemetry",
     "src/common/grpc",
     "src/common/grpc-expr",
+    "src/common/macro",
     "src/common/mem-prof",
     "src/common/meta",
+    "src/common/options",
     "src/common/plugins",
+    "src/common/pprof",
     "src/common/procedure",
     "src/common/procedure-test",
     "src/common/query",
@@ -29,7 +33,6 @@ members = [
     "src/common/telemetry",
     "src/common/test-util",
     "src/common/time",
-    "src/common/decimal",
     "src/common/version",
     "src/common/wal",
     "src/datanode",
@@ -37,6 +40,8 @@ members = [
     "src/file-engine",
     "src/flow",
     "src/frontend",
+    "src/index",
+    "src/log-query",
     "src/log-store",
     "src/meta-client",
     "src/meta-srv",
@@ -56,7 +61,6 @@ members = [
     "src/sql",
     "src/store-api",
     "src/table",
-    "src/index",
     "tests-fuzz",
     "tests-integration",
     "tests/runner",
@@ -64,7 +68,7 @@ members = [
 resolver = "2"
 
 [workspace.package]
-version = "0.9.1"
+version = "0.11.0"
 edition = "2021"
 license = "Apache-2.0"
 
@@ -77,6 +81,7 @@ clippy.readonly_write_lock = "allow"
 rust.unknown_lints = "deny"
 # Remove this after https://github.com/PyO3/pyo3/issues/4094
 rust.non_local_definitions = "allow"
+rust.unexpected_cfgs = { level = "warn", check-cfg = ['cfg(tokio_unstable)'] }
 
 [workspace.dependencies]
 # We turn off default-features for some dependencies here so the workspaces which inherit them can
@@ -89,7 +94,7 @@ aquamarine = "0.3"
 arrow = { version = "51.0.0", features = ["prettyprint"] }
 arrow-array = { version = "51.0.0", default-features = false, features = ["chrono-tz"] }
 arrow-flight = "51.0"
-arrow-ipc = { version = "51.0.0", default-features = false, features = ["lz4"] }
+arrow-ipc = { version = "51.0.0", default-features = false, features = ["lz4", "zstd"] }
 arrow-schema = { version = "51.0", features = ["serde"] }
 async-stream = "0.3"
 async-trait = "0.1"
@@ -98,7 +103,7 @@ base64 = "0.21"
 bigdecimal = "0.4.2"
 bitflags = "2.4.1"
 bytemuck = "1.12"
-bytes = { version = "1.5", features = ["serde"] }
+bytes = { version = "1.7", features = ["serde"] }
 chrono = { version = "0.4", features = ["serde"] }
 clap = { version = "4.4", features = ["derive"] }
 config = "0.13.0"
@@ -115,16 +120,18 @@ datafusion-sql = { git = "https://github.com/waynexia/arrow-datafusion.git", rev
 datafusion-substrait = { git = "https://github.com/waynexia/arrow-datafusion.git", rev = "7823ef2f63663907edab46af0d51359900f608d6" }
 derive_builder = "0.12"
 dotenv = "0.15"
-etcd-client = { version = "0.13" }
+etcd-client = "0.13"
 fst = "0.4.7"
 futures = "0.3"
 futures-util = "0.3"
-greptime-proto = { git = "https://github.com/GreptimeTeam/greptime-proto.git", rev = "c437b55725b7f5224fe9d46db21072b4a682ee4b" }
+greptime-proto = { git = "https://github.com/GreptimeTeam/greptime-proto.git", rev = "a875e976441188028353f7274a46a7e6e065c5d4" }
+hex = "0.4"
 humantime = "2.1"
 humantime-serde = "1.1"
 itertools = "0.10"
+jsonb = { git = "https://github.com/databendlabs/jsonb.git", rev = "8c8d2fc294a39f3ff08909d60f718639cfba3875", default-features = false }
 lazy_static = "1.4"
-meter-core = { git = "https://github.com/GreptimeTeam/greptime-meter.git", rev = "049171eb16cb4249d8099751a0c46750d1fe88e7" }
+meter-core = { git = "https://github.com/GreptimeTeam/greptime-meter.git", rev = "a10facb353b41460eeb98578868ebf19c2084fac" }
 mockall = "0.11.4"
 moka = "0.12"
 notify = "6.1"
@@ -134,33 +141,39 @@ opentelemetry-proto = { version = "0.5", features = [
     "gen-tonic",
     "metrics",
     "trace",
+    "with-serde",
+    "logs",
 ] }
+parking_lot = "0.12"
 parquet = { version = "51.0.0", default-features = false, features = ["arrow", "async", "object_store"] }
 paste = "1.0"
 pin-project = "1.0"
 prometheus = { version = "0.13.3", features = ["process"] }
-promql-parser = { version = "0.4" }
+promql-parser = { version = "0.4.3", features = ["ser"] }
 prost = "0.12"
 raft-engine = { version = "0.4.1", default-features = false }
 rand = "0.8"
+ratelimit = "0.9"
 regex = "1.8"
-regex-automata = { version = "0.4" }
+regex-automata = "0.4"
 reqwest = { version = "0.12", default-features = false, features = [
     "json",
     "rustls-tls-native-roots",
     "stream",
     "multipart",
 ] }
-rskafka = "0.5"
+rskafka = { git = "https://github.com/influxdata/rskafka.git", rev = "75535b5ad9bae4a5dbb582c82e44dfd81ec10105", features = [
+    "transport-tls",
+] }
 rstest = "0.21"
 rstest_reuse = "0.7"
 rust_decimal = "1.33"
 rustc-hash = "2.0"
-schemars = "0.8"
 serde = { version = "1.0", features = ["derive"] }
 serde_json = { version = "1.0", features = ["float_roundtrip"] }
 serde_with = "3"
-shadow-rs = "0.31"
+shadow-rs = "0.35"
+similar-asserts = "1.6.0"
 smallvec = { version = "1", features = ["serde"] }
 snafu = "0.8"
 sysinfo = "0.30"
@@ -170,12 +183,16 @@ sqlparser = { git = "https://github.com/GreptimeTeam/sqlparser-rs.git", rev = "5
 ] }
 strum = { version = "0.25", features = ["derive"] }
 tempfile = "3"
-tokio = { version = "1.36", features = ["full"] }
-tokio-stream = { version = "0.1" }
+tokio = { version = "1.40", features = ["full"] }
+tokio-postgres = "0.7"
+tokio-stream = "0.1"
 tokio-util = { version = "0.7", features = ["io-util", "compat"] }
 toml = "0.8.8"
 tonic = { version = "0.11", features = ["tls", "gzip", "zstd"] }
-tower = { version = "0.4" }
+tower = "0.4"
+tracing-appender = "0.2"
+tracing-subscriber = { version = "0.3", features = ["env-filter", "json", "fmt"] }
+typetag = "0.2"
 uuid = { version = "1.7", features = ["serde", "v4", "fast-rng"] }
 zstd = "0.13"
 
@@ -184,6 +201,7 @@ api = { path = "src/api" }
 auth = { path = "src/auth" }
 cache = { path = "src/cache" }
 catalog = { path = "src/catalog" }
+cli = { path = "src/cli" }
 client = { path = "src/client" }
 cmd = { path = "src/cmd", default-features = false }
 common-base = { path = "src/common/base" }
@@ -200,7 +218,9 @@ common-grpc-expr = { path = "src/common/grpc-expr" }
 common-macro = { path = "src/common/macro" }
 common-mem-prof = { path = "src/common/mem-prof" }
 common-meta = { path = "src/common/meta" }
+common-options = { path = "src/common/options" }
 common-plugins = { path = "src/common/plugins" }
+common-pprof = { path = "src/common/pprof" }
 common-procedure = { path = "src/common/procedure" }
 common-procedure-test = { path = "src/common/procedure-test" }
 common-query = { path = "src/common/query" }
@@ -238,16 +258,27 @@ store-api = { path = "src/store-api" }
 substrait = { path = "src/common/substrait" }
 table = { path = "src/table" }
 
+[patch.crates-io]
+# change all rustls dependencies to use our fork to default to `ring` to make it "just work"
+hyper-rustls = { git = "https://github.com/GreptimeTeam/hyper-rustls" }
+rustls = { git = "https://github.com/GreptimeTeam/rustls" }
+tokio-rustls = { git = "https://github.com/GreptimeTeam/tokio-rustls" }
+# This is commented, since we are not using aws-lc-sys, if we need to use it, we need to uncomment this line or use a release after this commit, or it wouldn't compile with gcc < 8.1
+# see https://github.com/aws/aws-lc-rs/pull/526
+# aws-lc-sys = { git ="https://github.com/aws/aws-lc-rs", rev = "556558441e3494af4b156ae95ebc07ebc2fd38aa" }
+# Apply a fix for pprof for unaligned pointer access
+pprof = { git = "https://github.com/GreptimeTeam/pprof-rs", rev = "1bd1e21" }
+
 [workspace.dependencies.meter-macros]
 git = "https://github.com/GreptimeTeam/greptime-meter.git"
-rev = "049171eb16cb4249d8099751a0c46750d1fe88e7"
+rev = "a10facb353b41460eeb98578868ebf19c2084fac"
 
 [profile.release]
 debug = 1
 
 [profile.nightly]
 inherits = "release"
-strip = true
+strip = "debuginfo"
 lto = "thin"
 debug = false
 incremental = false

Makefile

@@ -8,6 +8,7 @@ CARGO_BUILD_OPTS := --locked
 IMAGE_REGISTRY ?= docker.io
 IMAGE_NAMESPACE ?= greptime
 IMAGE_TAG ?= latest
+DEV_BUILDER_IMAGE_TAG ?= 2024-10-19-a5c00e85-20241024184445
 BUILDX_MULTI_PLATFORM_BUILD ?= false
 BUILDX_BUILDER_NAME ?= gtbuilder
 BASE_IMAGE ?= ubuntu
@@ -77,7 +78,7 @@ build: ## Build debug version greptime.
 build-by-dev-builder: ## Build greptime by dev-builder.
 	docker run --network=host \
 	-v ${PWD}:/greptimedb -v ${CARGO_REGISTRY_CACHE}:/root/.cargo/registry \
-	-w /greptimedb ${IMAGE_REGISTRY}/${IMAGE_NAMESPACE}/dev-builder-${BASE_IMAGE}:latest \
+	-w /greptimedb ${IMAGE_REGISTRY}/${IMAGE_NAMESPACE}/dev-builder-${BASE_IMAGE}:${DEV_BUILDER_IMAGE_TAG} \
 	make build \
 	CARGO_EXTENSION="${CARGO_EXTENSION}" \
 	CARGO_PROFILE=${CARGO_PROFILE} \
@@ -91,7 +92,7 @@ build-by-dev-builder: ## Build greptime by dev-builder.
 build-android-bin: ## Build greptime binary for android.
 	docker run --network=host \
 	-v ${PWD}:/greptimedb -v ${CARGO_REGISTRY_CACHE}:/root/.cargo/registry \
-	-w /greptimedb ${IMAGE_REGISTRY}/${IMAGE_NAMESPACE}/dev-builder-android:latest \
+	-w /greptimedb ${IMAGE_REGISTRY}/${IMAGE_NAMESPACE}/dev-builder-android:${DEV_BUILDER_IMAGE_TAG} \
 	make build \
 	CARGO_EXTENSION="ndk --platform 23 -t aarch64-linux-android" \
 	CARGO_PROFILE=release \
@@ -105,8 +106,8 @@ build-android-bin: ## Build greptime binary for android.
 strip-android-bin: build-android-bin ## Strip greptime binary for android.
 	docker run --network=host \
 	-v ${PWD}:/greptimedb \
-	-w /greptimedb ${IMAGE_REGISTRY}/${IMAGE_NAMESPACE}/dev-builder-android:latest \
-	bash -c '$${NDK_ROOT}/toolchains/llvm/prebuilt/linux-x86_64/bin/llvm-strip /greptimedb/target/aarch64-linux-android/release/greptime'
+	-w /greptimedb ${IMAGE_REGISTRY}/${IMAGE_NAMESPACE}/dev-builder-android:${DEV_BUILDER_IMAGE_TAG} \
+	bash -c '$${NDK_ROOT}/toolchains/llvm/prebuilt/linux-x86_64/bin/llvm-strip --strip-debug /greptimedb/target/aarch64-linux-android/release/greptime'
 
 .PHONY: clean
 clean: ## Clean the project.
@@ -145,7 +146,7 @@ dev-builder: multi-platform-buildx ## Build dev-builder image.
 	docker buildx build --builder ${BUILDX_BUILDER_NAME} \
 	--build-arg="RUST_TOOLCHAIN=${RUST_TOOLCHAIN}" \
 	-f docker/dev-builder/${BASE_IMAGE}/Dockerfile \
-	-t ${IMAGE_REGISTRY}/${IMAGE_NAMESPACE}/dev-builder-${BASE_IMAGE}:${IMAGE_TAG} ${BUILDX_MULTI_PLATFORM_BUILD_OPTS} .
+	-t ${IMAGE_REGISTRY}/${IMAGE_NAMESPACE}/dev-builder-${BASE_IMAGE}:${DEV_BUILDER_IMAGE_TAG} ${BUILDX_MULTI_PLATFORM_BUILD_OPTS} .
 
 .PHONY: multi-platform-buildx
 multi-platform-buildx: ## Create buildx multi-platform builder.
@@ -190,6 +191,7 @@ fix-clippy: ## Fix clippy violations.
 .PHONY: fmt-check
 fmt-check: ## Check code format.
 	cargo fmt --all -- --check
+	python3 scripts/check-snafu.py
 
 .PHONY: start-etcd
 start-etcd: ## Start single node etcd for testing purpose.
@@ -203,7 +205,7 @@ stop-etcd: ## Stop single node etcd for testing purpose.
 run-it-in-container: start-etcd ## Run integration tests in dev-builder.
 	docker run --network=host \
 	-v ${PWD}:/greptimedb -v ${CARGO_REGISTRY_CACHE}:/root/.cargo/registry -v /tmp:/tmp \
-	-w /greptimedb ${IMAGE_REGISTRY}/${IMAGE_NAMESPACE}/dev-builder-${BASE_IMAGE}:latest \
+	-w /greptimedb ${IMAGE_REGISTRY}/${IMAGE_NAMESPACE}/dev-builder-${BASE_IMAGE}:${DEV_BUILDER_IMAGE_TAG} \
 	make test sqlness-test BUILD_JOBS=${BUILD_JOBS}
 
 .PHONY: start-cluster
@@ -219,7 +221,7 @@ config-docs: ## Generate configuration documentation from toml files.
 	docker run --rm \
     -v ${PWD}:/greptimedb \
     -w /greptimedb/config \
-    toml2docs/toml2docs:v0.1.1 \
+    toml2docs/toml2docs:v0.1.3 \
     -p '##' \
     -t ./config-docs-template.md \
     -o ./config.md

README.md

@@ -6,7 +6,7 @@
   </picture>
 </p>
 
-<h2 align="center">Unified Time Series Database for Metrics, Logs, and Events</h2>
+<h2 align="center">Unified & Cost-Effective Time Series Database for Metrics, Logs, and Events</h2>
 
 <div align="center">
 <h3 align="center">
@@ -48,37 +48,51 @@
 </a>
 </div>
 
+- [Introduction](#introduction)
+- [**Features: Why GreptimeDB**](#why-greptimedb)
+- [Architecture](https://docs.greptime.com/contributor-guide/overview/#architecture)
+- [Try it for free](#try-greptimedb)
+- [Getting Started](#getting-started)
+- [Project Status](#project-status)
+- [Join the community](#community)
+  - [Contributing](#contributing)
+- [Tools & Extensions](#tools--extensions)
+- [License](#license)
+- [Acknowledgement](#acknowledgement)
+
 ## Introduction
 
-**GreptimeDB** is an open-source unified time-series database for **Metrics**, **Logs**, and **Events** (also **Traces** in plan). You can gain real-time insights from Edge to Cloud at any scale.
+**GreptimeDB** is an open-source unified & cost-effective time-series database for **Metrics**, **Logs**, and **Events** (also **Traces** in plan). You can gain real-time insights from Edge to Cloud at Any Scale.
 
 ## Why GreptimeDB
 
-Our core developers have been building time-series data platforms for years. Based on our best-practices, GreptimeDB is born to give you:
+Our core developers have been building time-series data platforms for years. Based on our best practices, GreptimeDB was born to give you:
 
-* **Unified all kinds of time series**
+* **Unified Processing of Metrics, Logs, and Events**
 
-  GreptimeDB treats all time series as contextual events with timestamp, and thus unifies the processing of metrics, logs, and events. It supports analyzing metrics, logs, and events with SQL and PromQL, and doing streaming with continuous aggregation.
+GreptimeDB unifies time series data processing by treating all data - whether metrics, logs, or events - as timestamped events with context. Users can analyze this data using either [SQL](https://docs.greptime.com/user-guide/query-data/sql) or [PromQL](https://docs.greptime.com/user-guide/query-data/promql) and leverage stream processing ([Flow](https://docs.greptime.com/user-guide/continuous-aggregation/overview)) to enable continuous aggregation. [Read more](https://docs.greptime.com/user-guide/concepts/data-model).
 
-* **Cloud-Edge collaboration**
+* **Cloud-native Distributed Database**
 
-  GreptimeDB can be deployed on ARM architecture-compatible Android/Linux systems as well as cloud environments from various vendors. Both sides run the same software, providing identical APIs and control planes, so your application can run at the edge or on the cloud without modification, and data synchronization also becomes extremely easy and efficient.
-
-* **Cloud-native distributed database**
-
-  By leveraging object storage (S3 and others), separating compute and storage, scaling stateless compute nodes arbitrarily, GreptimeDB implements seamless scalability. It also supports cross-cloud deployment with a built-in unified data access layer over different object storages.
+Built for [Kubernetes](https://docs.greptime.com/user-guide/deployments/deploy-on-kubernetes/greptimedb-operator-management). GreptimeDB achieves seamless scalability with its [cloud-native architecture](https://docs.greptime.com/user-guide/concepts/architecture) of separated compute and storage, built on object storage (AWS S3, Azure Blob Storage, etc.) while enabling cross-cloud deployment through a unified data access layer.
 
 * **Performance and Cost-effective**
 
-  Flexible indexing capabilities and distributed, parallel-processing query engine, tackling high cardinality issues down. Optimized columnar layout for handling time-series data; compacted, compressed, and stored on various storage backends, particularly cloud object storage with 50x cost efficiency.
+Written in pure Rust for superior performance and reliability. GreptimeDB features a distributed query engine with intelligent indexing to handle high cardinality data efficiently. Its optimized columnar storage achieves 50x cost efficiency on cloud object storage through advanced compression. [Benchmark reports](https://www.greptime.com/blogs/2024-09-09-report-summary).
 
-* **Compatible with InfluxDB, Prometheus and more protocols**
+* **Cloud-Edge Collaboration**
 
-  Widely adopted database protocols and APIs, including MySQL, PostgreSQL, and Prometheus Remote Storage, etc. [Read more](https://docs.greptime.com/user-guide/clients/overview).
+GreptimeDB seamlessly operates across cloud and edge (ARM/Android/Linux), providing consistent APIs and control plane for unified data management and efficient synchronization. [Learn how to run on Android](https://docs.greptime.com/user-guide/deployments/run-on-android/).
+
+* **Multi-protocol Ingestion, SQL & PromQL Ready**
+
+Widely adopted database protocols and APIs, including MySQL, PostgreSQL, InfluxDB, OpenTelemetry, Loki and Prometheus, etc.  Effortless Adoption & Seamless Migration. [Supported Protocols Overview](https://docs.greptime.com/user-guide/protocols/overview).
+
+For more detailed info please read  [Why GreptimeDB](https://docs.greptime.com/user-guide/concepts/why-greptimedb).
 
 ## Try GreptimeDB
 
-### 1. [GreptimePlay](https://greptime.com/playground)
+### 1. [Live Demo](https://greptime.com/playground)
 
 Try out the features of GreptimeDB right from your browser.
 
@@ -97,9 +111,18 @@ docker pull greptime/greptimedb
 Start a GreptimeDB container with:
 
 ```shell
-docker run --rm --name greptime --net=host greptime/greptimedb standalone start
+docker run -p 127.0.0.1:4000-4003:4000-4003 \
+  -v "$(pwd)/greptimedb:/tmp/greptimedb" \
+  --name greptime --rm \
+  greptime/greptimedb:latest standalone start \
+  --http-addr 0.0.0.0:4000 \
+  --rpc-addr 0.0.0.0:4001 \
+  --mysql-addr 0.0.0.0:4002 \
+  --postgres-addr 0.0.0.0:4003
 ```
 
+Access the dashboard via `http://localhost:4000/dashboard`.
+
 Read more about [Installation](https://docs.greptime.com/getting-started/installation/overview) on docs.
 
 ## Getting Started
@@ -129,7 +152,7 @@ Run a standalone server:
 cargo run -- standalone start
 ```
 
-## Extension
+## Tools & Extensions
 
 ### Dashboard
 
@@ -146,14 +169,19 @@ cargo run -- standalone start
 
 ### Grafana Dashboard
 
-Our official Grafana dashboard is available at [grafana](grafana/README.md) directory.
+Our official Grafana dashboard for monitoring GreptimeDB is available at [grafana](grafana/README.md) directory.
 
 ## Project Status
 
-The current version has not yet reached the standards for General Availability. 
-According to our Greptime 2024 Roadmap, we aim to achieve a production-level version with the release of v1.0 by the end of 2024. [Join Us](https://github.com/GreptimeTeam/greptimedb/issues/3412)
+GreptimeDB is currently in Beta. We are targeting GA (General Availability) with v1.0 release by Early 2025. 
 
-We welcome you to test and use GreptimeDB. Some users have already adopted it in their production environments. If you're interested in trying it out, please use the latest stable release available.
+While in Beta, GreptimeDB is already:
+
+* Being used in production by early adopters
+* Actively maintained with regular releases, [about version number](https://docs.greptime.com/nightly/reference/about-greptimedb-version)
+* Suitable for testing and evaluation
+
+For production use, we recommend using the latest stable release.
 
 ## Community
 
@@ -172,6 +200,13 @@ In addition, you may:
 - Connect us with [Linkedin](https://www.linkedin.com/company/greptime/)
 - Follow us on [Twitter](https://twitter.com/greptime)
 
+## Commercial Support
+
+If you are running GreptimeDB OSS in your organization, we offer additional
+enterprise add-ons, installation services, training, and consulting. [Contact
+us](https://greptime.com/contactus) and we will reach out to you with more
+detail of our commercial license.
+
 ## License
 
 GreptimeDB uses the [Apache License 2.0](https://apache.org/licenses/LICENSE-2.0.txt) to strike a balance between

config/config.md

@@ -14,7 +14,10 @@
 | --- | -----| ------- | ----------- |
 | `mode` | String | `standalone` | The running mode of the datanode. It can be `standalone` or `distributed`. |
 | `enable_telemetry` | Bool | `true` | Enable telemetry to collect anonymous usage data. |
-| `default_timezone` | String | `None` | The default timezone of the server. |
+| `default_timezone` | String | Unset | The default timezone of the server. |
+| `init_regions_in_background` | Bool | `false` | Initialize all regions in the background during the startup.<br/>By default, it provides services after all regions have been initialized. |
+| `init_regions_parallelism` | Integer | `16` | Parallelism of initializing regions. |
+| `max_concurrent_queries` | Integer | `0` | The maximum current queries allowed to be executed. Zero means unlimited. |
 | `runtime` | -- | -- | The runtime options. |
 | `runtime.global_rt_size` | Integer | `8` | The number of threads to execute the runtime for global read operations. |
 | `runtime.compact_rt_size` | Integer | `4` | The number of threads to execute the runtime for global write operations. |
@@ -27,8 +30,8 @@
 | `grpc.runtime_size` | Integer | `8` | The number of server worker threads. |
 | `grpc.tls` | -- | -- | gRPC server TLS options, see `mysql.tls` section. |
 | `grpc.tls.mode` | String | `disable` | TLS mode. |
-| `grpc.tls.cert_path` | String | `None` | Certificate file path. |
-| `grpc.tls.key_path` | String | `None` | Private key file path. |
+| `grpc.tls.cert_path` | String | Unset | Certificate file path. |
+| `grpc.tls.key_path` | String | Unset | Private key file path. |
 | `grpc.tls.watch` | Bool | `false` | Watch for Certificate and key file change and auto reload.<br/>For now, gRPC tls config does not support auto reload. |
 | `mysql` | -- | -- | MySQL server options. |
 | `mysql.enable` | Bool | `true` | Whether to enable. |
@@ -36,8 +39,8 @@
 | `mysql.runtime_size` | Integer | `2` | The number of server worker threads. |
 | `mysql.tls` | -- | -- | -- |
 | `mysql.tls.mode` | String | `disable` | TLS mode, refer to https://www.postgresql.org/docs/current/libpq-ssl.html<br/>- `disable` (default value)<br/>- `prefer`<br/>- `require`<br/>- `verify-ca`<br/>- `verify-full` |
-| `mysql.tls.cert_path` | String | `None` | Certificate file path. |
-| `mysql.tls.key_path` | String | `None` | Private key file path. |
+| `mysql.tls.cert_path` | String | Unset | Certificate file path. |
+| `mysql.tls.key_path` | String | Unset | Private key file path. |
 | `mysql.tls.watch` | Bool | `false` | Watch for Certificate and key file change and auto reload |
 | `postgres` | -- | -- | PostgresSQL server options. |
 | `postgres.enable` | Bool | `true` | Whether to enable |
@@ -45,8 +48,8 @@
 | `postgres.runtime_size` | Integer | `2` | The number of server worker threads. |
 | `postgres.tls` | -- | -- | PostgresSQL server TLS options, see `mysql.tls` section. |
 | `postgres.tls.mode` | String | `disable` | TLS mode. |
-| `postgres.tls.cert_path` | String | `None` | Certificate file path. |
-| `postgres.tls.key_path` | String | `None` | Private key file path. |
+| `postgres.tls.cert_path` | String | Unset | Certificate file path. |
+| `postgres.tls.key_path` | String | Unset | Private key file path. |
 | `postgres.tls.watch` | Bool | `false` | Watch for Certificate and key file change and auto reload |
 | `opentsdb` | -- | -- | OpenTSDB protocol options. |
 | `opentsdb.enable` | Bool | `true` | Whether to enable OpenTSDB put in HTTP API. |
@@ -57,7 +60,7 @@
 | `prom_store.with_metric_engine` | Bool | `true` | Whether to store the data from Prometheus remote write in metric engine. |
 | `wal` | -- | -- | The WAL options. |
 | `wal.provider` | String | `raft_engine` | The provider of the WAL.<br/>- `raft_engine`: the wal is stored in the local file system by raft-engine.<br/>- `kafka`: it's remote wal that data is stored in Kafka. |
-| `wal.dir` | String | `None` | The directory to store the WAL files.<br/>**It's only used when the provider is `raft_engine`**. |
+| `wal.dir` | String | Unset | The directory to store the WAL files.<br/>**It's only used when the provider is `raft_engine`**. |
 | `wal.file_size` | String | `256MB` | The size of the WAL segment file.<br/>**It's only used when the provider is `raft_engine`**. |
 | `wal.purge_threshold` | String | `4GB` | The threshold of the WAL size to trigger a flush.<br/>**It's only used when the provider is `raft_engine`**. |
 | `wal.purge_interval` | String | `10m` | The interval to trigger a flush.<br/>**It's only used when the provider is `raft_engine`**. |
@@ -66,13 +69,21 @@
 | `wal.enable_log_recycle` | Bool | `true` | Whether to reuse logically truncated log files.<br/>**It's only used when the provider is `raft_engine`**. |
 | `wal.prefill_log_files` | Bool | `false` | Whether to pre-create log files on start up.<br/>**It's only used when the provider is `raft_engine`**. |
 | `wal.sync_period` | String | `10s` | Duration for fsyncing log files.<br/>**It's only used when the provider is `raft_engine`**. |
+| `wal.recovery_parallelism` | Integer | `2` | Parallelism during WAL recovery. |
 | `wal.broker_endpoints` | Array | -- | The Kafka broker endpoints.<br/>**It's only used when the provider is `kafka`**. |
+| `wal.auto_create_topics` | Bool | `true` | Automatically create topics for WAL.<br/>Set to `true` to automatically create topics for WAL.<br/>Otherwise, use topics named `topic_name_prefix_[0..num_topics)` |
+| `wal.num_topics` | Integer | `64` | Number of topics.<br/>**It's only used when the provider is `kafka`**. |
+| `wal.selector_type` | String | `round_robin` | Topic selector type.<br/>Available selector types:<br/>- `round_robin` (default)<br/>**It's only used when the provider is `kafka`**. |
+| `wal.topic_name_prefix` | String | `greptimedb_wal_topic` | A Kafka topic is constructed by concatenating `topic_name_prefix` and `topic_id`.<br/>i.g., greptimedb_wal_topic_0, greptimedb_wal_topic_1.<br/>**It's only used when the provider is `kafka`**. |
+| `wal.replication_factor` | Integer | `1` | Expected number of replicas of each partition.<br/>**It's only used when the provider is `kafka`**. |
+| `wal.create_topic_timeout` | String | `30s` | Above which a topic creation operation will be cancelled.<br/>**It's only used when the provider is `kafka`**. |
 | `wal.max_batch_bytes` | String | `1MB` | The max size of a single producer batch.<br/>Warning: Kafka has a default limit of 1MB per message in a topic.<br/>**It's only used when the provider is `kafka`**. |
 | `wal.consumer_wait_timeout` | String | `100ms` | The consumer wait timeout.<br/>**It's only used when the provider is `kafka`**. |
 | `wal.backoff_init` | String | `500ms` | The initial backoff delay.<br/>**It's only used when the provider is `kafka`**. |
 | `wal.backoff_max` | String | `10s` | The maximum backoff delay.<br/>**It's only used when the provider is `kafka`**. |
 | `wal.backoff_base` | Integer | `2` | The exponential backoff rate, i.e. next backoff = base * current backoff.<br/>**It's only used when the provider is `kafka`**. |
 | `wal.backoff_deadline` | String | `5mins` | The deadline of retries.<br/>**It's only used when the provider is `kafka`**. |
+| `wal.overwrite_entry_start_id` | Bool | `false` | Ignore missing entries during read WAL.<br/>**It's only used when the provider is `kafka`**.<br/><br/>This option ensures that when Kafka messages are deleted, the system<br/>can still successfully replay memtable data without throwing an<br/>out-of-range error.<br/>However, enabling this option might lead to unexpected data loss,<br/>as the system will skip over missing entries instead of treating<br/>them as critical errors. |
 | `metadata_store` | -- | -- | Metadata storage options. |
 | `metadata_store.file_size` | String | `256MB` | Kv file size in bytes. |
 | `metadata_store.purge_threshold` | String | `4GB` | Kv purge threshold. |
@@ -82,21 +93,27 @@
 | `storage` | -- | -- | The data storage options. |
 | `storage.data_home` | String | `/tmp/greptimedb/` | The working home directory. |
 | `storage.type` | String | `File` | The storage type used to store the data.<br/>- `File`: the data is stored in the local file system.<br/>- `S3`: the data is stored in the S3 object storage.<br/>- `Gcs`: the data is stored in the Google Cloud Storage.<br/>- `Azblob`: the data is stored in the Azure Blob Storage.<br/>- `Oss`: the data is stored in the Aliyun OSS. |
-| `storage.cache_path` | String | `None` | Cache configuration for object storage such as 'S3' etc.<br/>The local file cache directory. |
-| `storage.cache_capacity` | String | `None` | The local file cache capacity in bytes. |
-| `storage.bucket` | String | `None` | The S3 bucket name.<br/>**It's only used when the storage type is `S3`, `Oss` and `Gcs`**. |
-| `storage.root` | String | `None` | The S3 data will be stored in the specified prefix, for example, `s3://${bucket}/${root}`.<br/>**It's only used when the storage type is `S3`, `Oss` and `Azblob`**. |
-| `storage.access_key_id` | String | `None` | The access key id of the aws account.<br/>It's **highly recommended** to use AWS IAM roles instead of hardcoding the access key id and secret key.<br/>**It's only used when the storage type is `S3` and `Oss`**. |
-| `storage.secret_access_key` | String | `None` | The secret access key of the aws account.<br/>It's **highly recommended** to use AWS IAM roles instead of hardcoding the access key id and secret key.<br/>**It's only used when the storage type is `S3`**. |
-| `storage.access_key_secret` | String | `None` | The secret access key of the aliyun account.<br/>**It's only used when the storage type is `Oss`**. |
-| `storage.account_name` | String | `None` | The account key of the azure account.<br/>**It's only used when the storage type is `Azblob`**. |
-| `storage.account_key` | String | `None` | The account key of the azure account.<br/>**It's only used when the storage type is `Azblob`**. |
-| `storage.scope` | String | `None` | The scope of the google cloud storage.<br/>**It's only used when the storage type is `Gcs`**. |
-| `storage.credential_path` | String | `None` | The credential path of the google cloud storage.<br/>**It's only used when the storage type is `Gcs`**. |
-| `storage.container` | String | `None` | The container of the azure account.<br/>**It's only used when the storage type is `Azblob`**. |
-| `storage.sas_token` | String | `None` | The sas token of the azure account.<br/>**It's only used when the storage type is `Azblob`**. |
-| `storage.endpoint` | String | `None` | The endpoint of the S3 service.<br/>**It's only used when the storage type is `S3`, `Oss`, `Gcs` and `Azblob`**. |
-| `storage.region` | String | `None` | The region of the S3 service.<br/>**It's only used when the storage type is `S3`, `Oss`, `Gcs` and `Azblob`**. |
+| `storage.cache_path` | String | Unset | Read cache configuration for object storage such as 'S3' etc, it's configured by default when using object storage. It is recommended to configure it when using object storage for better performance.<br/>A local file directory, defaults to `{data_home}/object_cache/read`. An empty string means disabling. |
+| `storage.cache_capacity` | String | Unset | The local file cache capacity in bytes. If your disk space is sufficient, it is recommended to set it larger. |
+| `storage.bucket` | String | Unset | The S3 bucket name.<br/>**It's only used when the storage type is `S3`, `Oss` and `Gcs`**. |
+| `storage.root` | String | Unset | The S3 data will be stored in the specified prefix, for example, `s3://${bucket}/${root}`.<br/>**It's only used when the storage type is `S3`, `Oss` and `Azblob`**. |
+| `storage.access_key_id` | String | Unset | The access key id of the aws account.<br/>It's **highly recommended** to use AWS IAM roles instead of hardcoding the access key id and secret key.<br/>**It's only used when the storage type is `S3` and `Oss`**. |
+| `storage.secret_access_key` | String | Unset | The secret access key of the aws account.<br/>It's **highly recommended** to use AWS IAM roles instead of hardcoding the access key id and secret key.<br/>**It's only used when the storage type is `S3`**. |
+| `storage.access_key_secret` | String | Unset | The secret access key of the aliyun account.<br/>**It's only used when the storage type is `Oss`**. |
+| `storage.account_name` | String | Unset | The account key of the azure account.<br/>**It's only used when the storage type is `Azblob`**. |
+| `storage.account_key` | String | Unset | The account key of the azure account.<br/>**It's only used when the storage type is `Azblob`**. |
+| `storage.scope` | String | Unset | The scope of the google cloud storage.<br/>**It's only used when the storage type is `Gcs`**. |
+| `storage.credential_path` | String | Unset | The credential path of the google cloud storage.<br/>**It's only used when the storage type is `Gcs`**. |
+| `storage.credential` | String | Unset | The credential of the google cloud storage.<br/>**It's only used when the storage type is `Gcs`**. |
+| `storage.container` | String | Unset | The container of the azure account.<br/>**It's only used when the storage type is `Azblob`**. |
+| `storage.sas_token` | String | Unset | The sas token of the azure account.<br/>**It's only used when the storage type is `Azblob`**. |
+| `storage.endpoint` | String | Unset | The endpoint of the S3 service.<br/>**It's only used when the storage type is `S3`, `Oss`, `Gcs` and `Azblob`**. |
+| `storage.region` | String | Unset | The region of the S3 service.<br/>**It's only used when the storage type is `S3`, `Oss`, `Gcs` and `Azblob`**. |
+| `storage.http_client` | -- | -- | The http client options to the storage.<br/>**It's only used when the storage type is `S3`, `Oss`, `Gcs` and `Azblob`**. |
+| `storage.http_client.pool_max_idle_per_host` | Integer | `1024` | The maximum idle connection per host allowed in the pool. |
+| `storage.http_client.connect_timeout` | String | `30s` | The timeout for only the connect phase of a http client. |
+| `storage.http_client.timeout` | String | `30s` | The total request timeout, applied from when the request starts connecting until the response body has finished.<br/>Also considered a total deadline. |
+| `storage.http_client.pool_idle_timeout` | String | `90s` | The timeout for idle sockets being kept-alive. |
 | `[[region_engine]]` | -- | -- | The region engine options. You can configure multiple region engines. |
 | `region_engine.mito` | -- | -- | The Mito engine options. |
 | `region_engine.mito.num_workers` | Integer | `8` | Number of region workers. |
@@ -104,21 +121,24 @@
 | `region_engine.mito.worker_request_batch_size` | Integer | `64` | Max batch size for a worker to handle requests. |
 | `region_engine.mito.manifest_checkpoint_distance` | Integer | `10` | Number of meta action updated to trigger a new checkpoint for the manifest. |
 | `region_engine.mito.compress_manifest` | Bool | `false` | Whether to compress manifest and checkpoint file by gzip (default false). |
-| `region_engine.mito.max_background_jobs` | Integer | `4` | Max number of running background jobs |
+| `region_engine.mito.max_background_flushes` | Integer | Auto | Max number of running background flush jobs (default: 1/2 of cpu cores). |
+| `region_engine.mito.max_background_compactions` | Integer | Auto | Max number of running background compaction jobs (default: 1/4 of cpu cores). |
+| `region_engine.mito.max_background_purges` | Integer | Auto | Max number of running background purge jobs (default: number of cpu cores). |
 | `region_engine.mito.auto_flush_interval` | String | `1h` | Interval to auto flush a region if it has not flushed yet. |
-| `region_engine.mito.global_write_buffer_size` | String | `1GB` | Global write buffer size for all regions. If not set, it's default to 1/8 of OS memory with a max limitation of 1GB. |
-| `region_engine.mito.global_write_buffer_reject_size` | String | `2GB` | Global write buffer size threshold to reject write requests. If not set, it's default to 2 times of `global_write_buffer_size` |
-| `region_engine.mito.sst_meta_cache_size` | String | `128MB` | Cache size for SST metadata. Setting it to 0 to disable the cache.<br/>If not set, it's default to 1/32 of OS memory with a max limitation of 128MB. |
-| `region_engine.mito.vector_cache_size` | String | `512MB` | Cache size for vectors and arrow arrays. Setting it to 0 to disable the cache.<br/>If not set, it's default to 1/16 of OS memory with a max limitation of 512MB. |
-| `region_engine.mito.page_cache_size` | String | `512MB` | Cache size for pages of SST row groups. Setting it to 0 to disable the cache.<br/>If not set, it's default to 1/16 of OS memory with a max limitation of 512MB. |
-| `region_engine.mito.enable_experimental_write_cache` | Bool | `false` | Whether to enable the experimental write cache. |
-| `region_engine.mito.experimental_write_cache_path` | String | `""` | File system path for write cache, defaults to `{data_home}/write_cache`. |
-| `region_engine.mito.experimental_write_cache_size` | String | `512MB` | Capacity for write cache. |
-| `region_engine.mito.experimental_write_cache_ttl` | String | `1h` | TTL for write cache. |
+| `region_engine.mito.global_write_buffer_size` | String | Auto | Global write buffer size for all regions. If not set, it's default to 1/8 of OS memory with a max limitation of 1GB. |
+| `region_engine.mito.global_write_buffer_reject_size` | String | Auto | Global write buffer size threshold to reject write requests. If not set, it's default to 2 times of `global_write_buffer_size`. |
+| `region_engine.mito.sst_meta_cache_size` | String | Auto | Cache size for SST metadata. Setting it to 0 to disable the cache.<br/>If not set, it's default to 1/32 of OS memory with a max limitation of 128MB. |
+| `region_engine.mito.vector_cache_size` | String | Auto | Cache size for vectors and arrow arrays. Setting it to 0 to disable the cache.<br/>If not set, it's default to 1/16 of OS memory with a max limitation of 512MB. |
+| `region_engine.mito.page_cache_size` | String | Auto | Cache size for pages of SST row groups. Setting it to 0 to disable the cache.<br/>If not set, it's default to 1/8 of OS memory. |
+| `region_engine.mito.selector_result_cache_size` | String | Auto | Cache size for time series selector (e.g. `last_value()`). Setting it to 0 to disable the cache.<br/>If not set, it's default to 1/16 of OS memory with a max limitation of 512MB. |
+| `region_engine.mito.enable_experimental_write_cache` | Bool | `false` | Whether to enable the experimental write cache, it's enabled by default when using object storage. It is recommended to enable it when using object storage for better performance. |
+| `region_engine.mito.experimental_write_cache_path` | String | `""` | File system path for write cache, defaults to `{data_home}/object_cache/write`. |
+| `region_engine.mito.experimental_write_cache_size` | String | `5GiB` | Capacity for write cache. If your disk space is sufficient, it is recommended to set it larger. |
+| `region_engine.mito.experimental_write_cache_ttl` | String | Unset | TTL for write cache. |
 | `region_engine.mito.sst_write_buffer_size` | String | `8MB` | Buffer size for SST writing. |
-| `region_engine.mito.scan_parallelism` | Integer | `0` | Parallelism to scan a region (default: 1/4 of cpu cores).<br/>- `0`: using the default value (1/4 of cpu cores).<br/>- `1`: scan in current thread.<br/>- `n`: scan in parallelism n. |
 | `region_engine.mito.parallel_scan_channel_size` | Integer | `32` | Capacity of the channel to send data from parallel scan tasks to the main task. |
 | `region_engine.mito.allow_stale_entries` | Bool | `false` | Whether to allow stale WAL entries read during replay. |
+| `region_engine.mito.min_compaction_interval` | String | `0m` | Minimum time interval between two compactions.<br/>To align with the old behavior, the default value is 0 (no restrictions). |
 | `region_engine.mito.index` | -- | -- | The options for index in Mito engine. |
 | `region_engine.mito.index.aux_path` | String | `""` | Auxiliary directory path for the index in filesystem, used to store intermediate files for<br/>creating the index and staging files for searching the index, defaults to `{data_home}/index_intermediate`.<br/>The default name for this directory is `index_intermediate` for backward compatibility.<br/><br/>This path contains two subdirectories:<br/>- `__intm`: for storing intermediate files used during creating index.<br/>- `staging`: for storing staging files used during searching index. |
 | `region_engine.mito.index.staging_size` | String | `2GB` | The max capacity of the staging directory. |
@@ -142,23 +162,29 @@
 | `region_engine.mito.memtable.fork_dictionary_bytes` | String | `1GiB` | Max dictionary bytes.<br/>Only available for `partition_tree` memtable. |
 | `region_engine.file` | -- | -- | Enable the file engine. |
 | `logging` | -- | -- | The logging options. |
-| `logging.dir` | String | `/tmp/greptimedb/logs` | The directory to store the log files. |
-| `logging.level` | String | `None` | The log level. Can be `info`/`debug`/`warn`/`error`. |
+| `logging.dir` | String | `/tmp/greptimedb/logs` | The directory to store the log files. If set to empty, logs will not be written to files. |
+| `logging.level` | String | Unset | The log level. Can be `info`/`debug`/`warn`/`error`. |
 | `logging.enable_otlp_tracing` | Bool | `false` | Enable OTLP tracing. |
 | `logging.otlp_endpoint` | String | `http://localhost:4317` | The OTLP tracing endpoint. |
 | `logging.append_stdout` | Bool | `true` | Whether to append logs to stdout. |
+| `logging.log_format` | String | `text` | The log format. Can be `text`/`json`. |
+| `logging.max_log_files` | Integer | `720` | The maximum amount of log files. |
 | `logging.tracing_sample_ratio` | -- | -- | The percentage of tracing will be sampled and exported.<br/>Valid range `[0, 1]`, 1 means all traces are sampled, 0 means all traces are not sampled, the default value is 1.<br/>ratio > 1 are treated as 1. Fractions < 0 are treated as 0 |
 | `logging.tracing_sample_ratio.default_ratio` | Float | `1.0` | -- |
+| `logging.slow_query` | -- | -- | The slow query log options. |
+| `logging.slow_query.enable` | Bool | `false` | Whether to enable slow query log. |
+| `logging.slow_query.threshold` | String | Unset | The threshold of slow query. |
+| `logging.slow_query.sample_ratio` | Float | Unset | The sampling ratio of slow query log. The value should be in the range of (0, 1]. |
 | `export_metrics` | -- | -- | The datanode can export its metrics and send to Prometheus compatible service (e.g. send to `greptimedb` itself) from remote-write API.<br/>This is only used for `greptimedb` to export its own metrics internally. It's different from prometheus scrape. |
 | `export_metrics.enable` | Bool | `false` | whether enable export metrics. |
 | `export_metrics.write_interval` | String | `30s` | The interval of export metrics. |
-| `export_metrics.self_import` | -- | -- | For `standalone` mode, `self_import` is recommend to collect metrics generated by itself |
-| `export_metrics.self_import.db` | String | `None` | -- |
+| `export_metrics.self_import` | -- | -- | For `standalone` mode, `self_import` is recommended to collect metrics generated by itself<br/>You must create the database before enabling it. |
+| `export_metrics.self_import.db` | String | Unset | -- |
 | `export_metrics.remote_write` | -- | -- | -- |
-| `export_metrics.remote_write.url` | String | `""` | The url the metrics send to. The url example can be: `http://127.0.0.1:4000/v1/prometheus/write?db=information_schema`. |
+| `export_metrics.remote_write.url` | String | `""` | The url the metrics send to. The url example can be: `http://127.0.0.1:4000/v1/prometheus/write?db=greptime_metrics`. |
 | `export_metrics.remote_write.headers` | InlineTable | -- | HTTP headers of Prometheus remote-write carry. |
 | `tracing` | -- | -- | The tracing options. Only effect when compiled with `tokio-console` feature. |
-| `tracing.tokio_console_addr` | String | `None` | The tokio console address. |
+| `tracing.tokio_console_addr` | String | Unset | The tokio console address. |
 
 
 ## Distributed Mode
@@ -167,7 +193,7 @@
 
 | Key | Type | Default | Descriptions |
 | --- | -----| ------- | ----------- |
-| `default_timezone` | String | `None` | The default timezone of the server. |
+| `default_timezone` | String | Unset | The default timezone of the server. |
 | `runtime` | -- | -- | The runtime options. |
 | `runtime.global_rt_size` | Integer | `8` | The number of threads to execute the runtime for global read operations. |
 | `runtime.compact_rt_size` | Integer | `4` | The number of threads to execute the runtime for global write operations. |
@@ -184,8 +210,8 @@
 | `grpc.runtime_size` | Integer | `8` | The number of server worker threads. |
 | `grpc.tls` | -- | -- | gRPC server TLS options, see `mysql.tls` section. |
 | `grpc.tls.mode` | String | `disable` | TLS mode. |
-| `grpc.tls.cert_path` | String | `None` | Certificate file path. |
-| `grpc.tls.key_path` | String | `None` | Private key file path. |
+| `grpc.tls.cert_path` | String | Unset | Certificate file path. |
+| `grpc.tls.key_path` | String | Unset | Private key file path. |
 | `grpc.tls.watch` | Bool | `false` | Watch for Certificate and key file change and auto reload.<br/>For now, gRPC tls config does not support auto reload. |
 | `mysql` | -- | -- | MySQL server options. |
 | `mysql.enable` | Bool | `true` | Whether to enable. |
@@ -193,8 +219,8 @@
 | `mysql.runtime_size` | Integer | `2` | The number of server worker threads. |
 | `mysql.tls` | -- | -- | -- |
 | `mysql.tls.mode` | String | `disable` | TLS mode, refer to https://www.postgresql.org/docs/current/libpq-ssl.html<br/>- `disable` (default value)<br/>- `prefer`<br/>- `require`<br/>- `verify-ca`<br/>- `verify-full` |
-| `mysql.tls.cert_path` | String | `None` | Certificate file path. |
-| `mysql.tls.key_path` | String | `None` | Private key file path. |
+| `mysql.tls.cert_path` | String | Unset | Certificate file path. |
+| `mysql.tls.key_path` | String | Unset | Private key file path. |
 | `mysql.tls.watch` | Bool | `false` | Watch for Certificate and key file change and auto reload |
 | `postgres` | -- | -- | PostgresSQL server options. |
 | `postgres.enable` | Bool | `true` | Whether to enable |
@@ -202,8 +228,8 @@
 | `postgres.runtime_size` | Integer | `2` | The number of server worker threads. |
 | `postgres.tls` | -- | -- | PostgresSQL server TLS options, see `mysql.tls` section. |
 | `postgres.tls.mode` | String | `disable` | TLS mode. |
-| `postgres.tls.cert_path` | String | `None` | Certificate file path. |
-| `postgres.tls.key_path` | String | `None` | Private key file path. |
+| `postgres.tls.cert_path` | String | Unset | Certificate file path. |
+| `postgres.tls.key_path` | String | Unset | Private key file path. |
 | `postgres.tls.watch` | Bool | `false` | Watch for Certificate and key file change and auto reload |
 | `opentsdb` | -- | -- | OpenTSDB protocol options. |
 | `opentsdb.enable` | Bool | `true` | Whether to enable OpenTSDB put in HTTP API. |
@@ -227,23 +253,29 @@
 | `datanode.client.connect_timeout` | String | `10s` | -- |
 | `datanode.client.tcp_nodelay` | Bool | `true` | -- |
 | `logging` | -- | -- | The logging options. |
-| `logging.dir` | String | `/tmp/greptimedb/logs` | The directory to store the log files. |
-| `logging.level` | String | `None` | The log level. Can be `info`/`debug`/`warn`/`error`. |
+| `logging.dir` | String | `/tmp/greptimedb/logs` | The directory to store the log files. If set to empty, logs will not be written to files. |
+| `logging.level` | String | Unset | The log level. Can be `info`/`debug`/`warn`/`error`. |
 | `logging.enable_otlp_tracing` | Bool | `false` | Enable OTLP tracing. |
 | `logging.otlp_endpoint` | String | `http://localhost:4317` | The OTLP tracing endpoint. |
 | `logging.append_stdout` | Bool | `true` | Whether to append logs to stdout. |
+| `logging.log_format` | String | `text` | The log format. Can be `text`/`json`. |
+| `logging.max_log_files` | Integer | `720` | The maximum amount of log files. |
 | `logging.tracing_sample_ratio` | -- | -- | The percentage of tracing will be sampled and exported.<br/>Valid range `[0, 1]`, 1 means all traces are sampled, 0 means all traces are not sampled, the default value is 1.<br/>ratio > 1 are treated as 1. Fractions < 0 are treated as 0 |
 | `logging.tracing_sample_ratio.default_ratio` | Float | `1.0` | -- |
+| `logging.slow_query` | -- | -- | The slow query log options. |
+| `logging.slow_query.enable` | Bool | `false` | Whether to enable slow query log. |
+| `logging.slow_query.threshold` | String | Unset | The threshold of slow query. |
+| `logging.slow_query.sample_ratio` | Float | Unset | The sampling ratio of slow query log. The value should be in the range of (0, 1]. |
 | `export_metrics` | -- | -- | The datanode can export its metrics and send to Prometheus compatible service (e.g. send to `greptimedb` itself) from remote-write API.<br/>This is only used for `greptimedb` to export its own metrics internally. It's different from prometheus scrape. |
 | `export_metrics.enable` | Bool | `false` | whether enable export metrics. |
 | `export_metrics.write_interval` | String | `30s` | The interval of export metrics. |
-| `export_metrics.self_import` | -- | -- | For `standalone` mode, `self_import` is recommend to collect metrics generated by itself |
-| `export_metrics.self_import.db` | String | `None` | -- |
+| `export_metrics.self_import` | -- | -- | For `standalone` mode, `self_import` is recommend to collect metrics generated by itself<br/>You must create the database before enabling it. |
+| `export_metrics.self_import.db` | String | Unset | -- |
 | `export_metrics.remote_write` | -- | -- | -- |
-| `export_metrics.remote_write.url` | String | `""` | The url the metrics send to. The url example can be: `http://127.0.0.1:4000/v1/prometheus/write?db=information_schema`. |
+| `export_metrics.remote_write.url` | String | `""` | The url the metrics send to. The url example can be: `http://127.0.0.1:4000/v1/prometheus/write?db=greptime_metrics`. |
 | `export_metrics.remote_write.headers` | InlineTable | -- | HTTP headers of Prometheus remote-write carry. |
 | `tracing` | -- | -- | The tracing options. Only effect when compiled with `tokio-console` feature. |
-| `tracing.tokio_console_addr` | String | `None` | The tokio console address. |
+| `tracing.tokio_console_addr` | String | Unset | The tokio console address. |
 
 
 ### Metasrv
@@ -253,12 +285,13 @@
 | `data_home` | String | `/tmp/metasrv/` | The working home directory. |
 | `bind_addr` | String | `127.0.0.1:3002` | The bind address of metasrv. |
 | `server_addr` | String | `127.0.0.1:3002` | The communication server address for frontend and datanode to connect to metasrv,  "127.0.0.1:3002" by default for localhost. |
-| `store_addr` | String | `127.0.0.1:2379` | Etcd server address. |
+| `store_addrs` | Array | -- | Store server address default to etcd store. |
 | `selector` | String | `round_robin` | Datanode selector type.<br/>- `round_robin` (default value)<br/>- `lease_based`<br/>- `load_based`<br/>For details, please see "https://docs.greptime.com/developer-guide/metasrv/selector". |
 | `use_memory_store` | Bool | `false` | Store data in memory. |
 | `enable_telemetry` | Bool | `true` | Whether to enable greptimedb telemetry. |
 | `store_key_prefix` | String | `""` | If it's not empty, the metasrv will store all data with this key prefix. |
 | `enable_region_failover` | Bool | `false` | Whether to enable region failover.<br/>This feature is only available on GreptimeDB running on cluster mode and<br/>- Using Remote WAL<br/>- Using shared storage (e.g., s3). |
+| `backend` | String | `EtcdStore` | The datastore for meta server. |
 | `runtime` | -- | -- | The runtime options. |
 | `runtime.global_rt_size` | Integer | `8` | The number of threads to execute the runtime for global read operations. |
 | `runtime.compact_rt_size` | Integer | `4` | The number of threads to execute the runtime for global write operations. |
@@ -279,9 +312,10 @@
 | `wal` | -- | -- | -- |
 | `wal.provider` | String | `raft_engine` | -- |
 | `wal.broker_endpoints` | Array | -- | The broker endpoints of the Kafka cluster. |
-| `wal.num_topics` | Integer | `64` | Number of topics to be created upon start. |
+| `wal.auto_create_topics` | Bool | `true` | Automatically create topics for WAL.<br/>Set to `true` to automatically create topics for WAL.<br/>Otherwise, use topics named `topic_name_prefix_[0..num_topics)` |
+| `wal.num_topics` | Integer | `64` | Number of topics. |
 | `wal.selector_type` | String | `round_robin` | Topic selector type.<br/>Available selector types:<br/>- `round_robin` (default) |
-| `wal.topic_name_prefix` | String | `greptimedb_wal_topic` | A Kafka topic is constructed by concatenating `topic_name_prefix` and `topic_id`. |
+| `wal.topic_name_prefix` | String | `greptimedb_wal_topic` | A Kafka topic is constructed by concatenating `topic_name_prefix` and `topic_id`.<br/>i.g., greptimedb_wal_topic_0, greptimedb_wal_topic_1. |
 | `wal.replication_factor` | Integer | `1` | Expected number of replicas of each partition. |
 | `wal.create_topic_timeout` | String | `30s` | Above which a topic creation operation will be cancelled. |
 | `wal.backoff_init` | String | `500ms` | The initial backoff for kafka clients. |
@@ -289,23 +323,29 @@
 | `wal.backoff_base` | Integer | `2` | Exponential backoff rate, i.e. next backoff = base * current backoff. |
 | `wal.backoff_deadline` | String | `5mins` | Stop reconnecting if the total wait time reaches the deadline. If this config is missing, the reconnecting won't terminate. |
 | `logging` | -- | -- | The logging options. |
-| `logging.dir` | String | `/tmp/greptimedb/logs` | The directory to store the log files. |
-| `logging.level` | String | `None` | The log level. Can be `info`/`debug`/`warn`/`error`. |
+| `logging.dir` | String | `/tmp/greptimedb/logs` | The directory to store the log files. If set to empty, logs will not be written to files. |
+| `logging.level` | String | Unset | The log level. Can be `info`/`debug`/`warn`/`error`. |
 | `logging.enable_otlp_tracing` | Bool | `false` | Enable OTLP tracing. |
 | `logging.otlp_endpoint` | String | `http://localhost:4317` | The OTLP tracing endpoint. |
 | `logging.append_stdout` | Bool | `true` | Whether to append logs to stdout. |
+| `logging.log_format` | String | `text` | The log format. Can be `text`/`json`. |
+| `logging.max_log_files` | Integer | `720` | The maximum amount of log files. |
 | `logging.tracing_sample_ratio` | -- | -- | The percentage of tracing will be sampled and exported.<br/>Valid range `[0, 1]`, 1 means all traces are sampled, 0 means all traces are not sampled, the default value is 1.<br/>ratio > 1 are treated as 1. Fractions < 0 are treated as 0 |
 | `logging.tracing_sample_ratio.default_ratio` | Float | `1.0` | -- |
+| `logging.slow_query` | -- | -- | The slow query log options. |
+| `logging.slow_query.enable` | Bool | `false` | Whether to enable slow query log. |
+| `logging.slow_query.threshold` | String | Unset | The threshold of slow query. |
+| `logging.slow_query.sample_ratio` | Float | Unset | The sampling ratio of slow query log. The value should be in the range of (0, 1]. |
 | `export_metrics` | -- | -- | The datanode can export its metrics and send to Prometheus compatible service (e.g. send to `greptimedb` itself) from remote-write API.<br/>This is only used for `greptimedb` to export its own metrics internally. It's different from prometheus scrape. |
 | `export_metrics.enable` | Bool | `false` | whether enable export metrics. |
 | `export_metrics.write_interval` | String | `30s` | The interval of export metrics. |
-| `export_metrics.self_import` | -- | -- | For `standalone` mode, `self_import` is recommend to collect metrics generated by itself |
-| `export_metrics.self_import.db` | String | `None` | -- |
+| `export_metrics.self_import` | -- | -- | For `standalone` mode, `self_import` is recommend to collect metrics generated by itself<br/>You must create the database before enabling it. |
+| `export_metrics.self_import.db` | String | Unset | -- |
 | `export_metrics.remote_write` | -- | -- | -- |
-| `export_metrics.remote_write.url` | String | `""` | The url the metrics send to. The url example can be: `http://127.0.0.1:4000/v1/prometheus/write?db=information_schema`. |
+| `export_metrics.remote_write.url` | String | `""` | The url the metrics send to. The url example can be: `http://127.0.0.1:4000/v1/prometheus/write?db=greptime_metrics`. |
 | `export_metrics.remote_write.headers` | InlineTable | -- | HTTP headers of Prometheus remote-write carry. |
 | `tracing` | -- | -- | The tracing options. Only effect when compiled with `tokio-console` feature. |
-| `tracing.tokio_console_addr` | String | `None` | The tokio console address. |
+| `tracing.tokio_console_addr` | String | Unset | The tokio console address. |
 
 
 ### Datanode
@@ -313,16 +353,21 @@
 | Key | Type | Default | Descriptions |
 | --- | -----| ------- | ----------- |
 | `mode` | String | `standalone` | The running mode of the datanode. It can be `standalone` or `distributed`. |
-| `node_id` | Integer | `None` | The datanode identifier and should be unique in the cluster. |
+| `node_id` | Integer | Unset | The datanode identifier and should be unique in the cluster. |
 | `require_lease_before_startup` | Bool | `false` | Start services after regions have obtained leases.<br/>It will block the datanode start if it can't receive leases in the heartbeat from metasrv. |
 | `init_regions_in_background` | Bool | `false` | Initialize all regions in the background during the startup.<br/>By default, it provides services after all regions have been initialized. |
 | `enable_telemetry` | Bool | `true` | Enable telemetry to collect anonymous usage data. |
 | `init_regions_parallelism` | Integer | `16` | Parallelism of initializing regions. |
-| `rpc_addr` | String | `None` | Deprecated, use `grpc.addr` instead. |
-| `rpc_hostname` | String | `None` | Deprecated, use `grpc.hostname` instead. |
-| `rpc_runtime_size` | Integer | `None` | Deprecated, use `grpc.runtime_size` instead. |
-| `rpc_max_recv_message_size` | String | `None` | Deprecated, use `grpc.rpc_max_recv_message_size` instead. |
-| `rpc_max_send_message_size` | String | `None` | Deprecated, use `grpc.rpc_max_send_message_size` instead. |
+| `max_concurrent_queries` | Integer | `0` | The maximum current queries allowed to be executed. Zero means unlimited. |
+| `rpc_addr` | String | Unset | Deprecated, use `grpc.addr` instead. |
+| `rpc_hostname` | String | Unset | Deprecated, use `grpc.hostname` instead. |
+| `rpc_runtime_size` | Integer | Unset | Deprecated, use `grpc.runtime_size` instead. |
+| `rpc_max_recv_message_size` | String | Unset | Deprecated, use `grpc.rpc_max_recv_message_size` instead. |
+| `rpc_max_send_message_size` | String | Unset | Deprecated, use `grpc.rpc_max_send_message_size` instead. |
+| `http` | -- | -- | The HTTP server options. |
+| `http.addr` | String | `127.0.0.1:4000` | The address to bind the HTTP server. |
+| `http.timeout` | String | `30s` | HTTP request timeout. Set to 0 to disable timeout. |
+| `http.body_limit` | String | `64MB` | HTTP request body limit.<br/>The following units are supported: `B`, `KB`, `KiB`, `MB`, `MiB`, `GB`, `GiB`, `TB`, `TiB`, `PB`, `PiB`.<br/>Set to 0 to disable limit. |
 | `grpc` | -- | -- | The gRPC server options. |
 | `grpc.addr` | String | `127.0.0.1:3001` | The address to bind the gRPC server. |
 | `grpc.hostname` | String | `127.0.0.1` | The hostname advertised to the metasrv,<br/>and used for connections from outside the host |
@@ -331,8 +376,8 @@
 | `grpc.max_send_message_size` | String | `512MB` | The maximum send message size for gRPC server. |
 | `grpc.tls` | -- | -- | gRPC server TLS options, see `mysql.tls` section. |
 | `grpc.tls.mode` | String | `disable` | TLS mode. |
-| `grpc.tls.cert_path` | String | `None` | Certificate file path. |
-| `grpc.tls.key_path` | String | `None` | Private key file path. |
+| `grpc.tls.cert_path` | String | Unset | Certificate file path. |
+| `grpc.tls.key_path` | String | Unset | Private key file path. |
 | `grpc.tls.watch` | Bool | `false` | Watch for Certificate and key file change and auto reload.<br/>For now, gRPC tls config does not support auto reload. |
 | `runtime` | -- | -- | The runtime options. |
 | `runtime.global_rt_size` | Integer | `8` | The number of threads to execute the runtime for global read operations. |
@@ -352,7 +397,7 @@
 | `meta_client.metadata_cache_tti` | String | `5m` | -- |
 | `wal` | -- | -- | The WAL options. |
 | `wal.provider` | String | `raft_engine` | The provider of the WAL.<br/>- `raft_engine`: the wal is stored in the local file system by raft-engine.<br/>- `kafka`: it's remote wal that data is stored in Kafka. |
-| `wal.dir` | String | `None` | The directory to store the WAL files.<br/>**It's only used when the provider is `raft_engine`**. |
+| `wal.dir` | String | Unset | The directory to store the WAL files.<br/>**It's only used when the provider is `raft_engine`**. |
 | `wal.file_size` | String | `256MB` | The size of the WAL segment file.<br/>**It's only used when the provider is `raft_engine`**. |
 | `wal.purge_threshold` | String | `4GB` | The threshold of the WAL size to trigger a flush.<br/>**It's only used when the provider is `raft_engine`**. |
 | `wal.purge_interval` | String | `10m` | The interval to trigger a flush.<br/>**It's only used when the provider is `raft_engine`**. |
@@ -361,6 +406,7 @@
 | `wal.enable_log_recycle` | Bool | `true` | Whether to reuse logically truncated log files.<br/>**It's only used when the provider is `raft_engine`**. |
 | `wal.prefill_log_files` | Bool | `false` | Whether to pre-create log files on start up.<br/>**It's only used when the provider is `raft_engine`**. |
 | `wal.sync_period` | String | `10s` | Duration for fsyncing log files.<br/>**It's only used when the provider is `raft_engine`**. |
+| `wal.recovery_parallelism` | Integer | `2` | Parallelism during WAL recovery. |
 | `wal.broker_endpoints` | Array | -- | The Kafka broker endpoints.<br/>**It's only used when the provider is `kafka`**. |
 | `wal.max_batch_bytes` | String | `1MB` | The max size of a single producer batch.<br/>Warning: Kafka has a default limit of 1MB per message in a topic.<br/>**It's only used when the provider is `kafka`**. |
 | `wal.consumer_wait_timeout` | String | `100ms` | The consumer wait timeout.<br/>**It's only used when the provider is `kafka`**. |
@@ -368,24 +414,33 @@
 | `wal.backoff_max` | String | `10s` | The maximum backoff delay.<br/>**It's only used when the provider is `kafka`**. |
 | `wal.backoff_base` | Integer | `2` | The exponential backoff rate, i.e. next backoff = base * current backoff.<br/>**It's only used when the provider is `kafka`**. |
 | `wal.backoff_deadline` | String | `5mins` | The deadline of retries.<br/>**It's only used when the provider is `kafka`**. |
+| `wal.create_index` | Bool | `true` | Whether to enable WAL index creation.<br/>**It's only used when the provider is `kafka`**. |
+| `wal.dump_index_interval` | String | `60s` | The interval for dumping WAL indexes.<br/>**It's only used when the provider is `kafka`**. |
+| `wal.overwrite_entry_start_id` | Bool | `false` | Ignore missing entries during read WAL.<br/>**It's only used when the provider is `kafka`**.<br/><br/>This option ensures that when Kafka messages are deleted, the system<br/>can still successfully replay memtable data without throwing an<br/>out-of-range error.<br/>However, enabling this option might lead to unexpected data loss,<br/>as the system will skip over missing entries instead of treating<br/>them as critical errors. |
 | `storage` | -- | -- | The data storage options. |
 | `storage.data_home` | String | `/tmp/greptimedb/` | The working home directory. |
 | `storage.type` | String | `File` | The storage type used to store the data.<br/>- `File`: the data is stored in the local file system.<br/>- `S3`: the data is stored in the S3 object storage.<br/>- `Gcs`: the data is stored in the Google Cloud Storage.<br/>- `Azblob`: the data is stored in the Azure Blob Storage.<br/>- `Oss`: the data is stored in the Aliyun OSS. |
-| `storage.cache_path` | String | `None` | Cache configuration for object storage such as 'S3' etc.<br/>The local file cache directory. |
-| `storage.cache_capacity` | String | `None` | The local file cache capacity in bytes. |
-| `storage.bucket` | String | `None` | The S3 bucket name.<br/>**It's only used when the storage type is `S3`, `Oss` and `Gcs`**. |
-| `storage.root` | String | `None` | The S3 data will be stored in the specified prefix, for example, `s3://${bucket}/${root}`.<br/>**It's only used when the storage type is `S3`, `Oss` and `Azblob`**. |
-| `storage.access_key_id` | String | `None` | The access key id of the aws account.<br/>It's **highly recommended** to use AWS IAM roles instead of hardcoding the access key id and secret key.<br/>**It's only used when the storage type is `S3` and `Oss`**. |
-| `storage.secret_access_key` | String | `None` | The secret access key of the aws account.<br/>It's **highly recommended** to use AWS IAM roles instead of hardcoding the access key id and secret key.<br/>**It's only used when the storage type is `S3`**. |
-| `storage.access_key_secret` | String | `None` | The secret access key of the aliyun account.<br/>**It's only used when the storage type is `Oss`**. |
-| `storage.account_name` | String | `None` | The account key of the azure account.<br/>**It's only used when the storage type is `Azblob`**. |
-| `storage.account_key` | String | `None` | The account key of the azure account.<br/>**It's only used when the storage type is `Azblob`**. |
-| `storage.scope` | String | `None` | The scope of the google cloud storage.<br/>**It's only used when the storage type is `Gcs`**. |
-| `storage.credential_path` | String | `None` | The credential path of the google cloud storage.<br/>**It's only used when the storage type is `Gcs`**. |
-| `storage.container` | String | `None` | The container of the azure account.<br/>**It's only used when the storage type is `Azblob`**. |
-| `storage.sas_token` | String | `None` | The sas token of the azure account.<br/>**It's only used when the storage type is `Azblob`**. |
-| `storage.endpoint` | String | `None` | The endpoint of the S3 service.<br/>**It's only used when the storage type is `S3`, `Oss`, `Gcs` and `Azblob`**. |
-| `storage.region` | String | `None` | The region of the S3 service.<br/>**It's only used when the storage type is `S3`, `Oss`, `Gcs` and `Azblob`**. |
+| `storage.cache_path` | String | Unset | Read cache configuration for object storage such as 'S3' etc, it's configured by default when using object storage. It is recommended to configure it when using object storage for better performance.<br/>A local file directory, defaults to `{data_home}/object_cache/read`. An empty string means disabling. |
+| `storage.cache_capacity` | String | Unset | The local file cache capacity in bytes. If your disk space is sufficient, it is recommended to set it larger. |
+| `storage.bucket` | String | Unset | The S3 bucket name.<br/>**It's only used when the storage type is `S3`, `Oss` and `Gcs`**. |
+| `storage.root` | String | Unset | The S3 data will be stored in the specified prefix, for example, `s3://${bucket}/${root}`.<br/>**It's only used when the storage type is `S3`, `Oss` and `Azblob`**. |
+| `storage.access_key_id` | String | Unset | The access key id of the aws account.<br/>It's **highly recommended** to use AWS IAM roles instead of hardcoding the access key id and secret key.<br/>**It's only used when the storage type is `S3` and `Oss`**. |
+| `storage.secret_access_key` | String | Unset | The secret access key of the aws account.<br/>It's **highly recommended** to use AWS IAM roles instead of hardcoding the access key id and secret key.<br/>**It's only used when the storage type is `S3`**. |
+| `storage.access_key_secret` | String | Unset | The secret access key of the aliyun account.<br/>**It's only used when the storage type is `Oss`**. |
+| `storage.account_name` | String | Unset | The account key of the azure account.<br/>**It's only used when the storage type is `Azblob`**. |
+| `storage.account_key` | String | Unset | The account key of the azure account.<br/>**It's only used when the storage type is `Azblob`**. |
+| `storage.scope` | String | Unset | The scope of the google cloud storage.<br/>**It's only used when the storage type is `Gcs`**. |
+| `storage.credential_path` | String | Unset | The credential path of the google cloud storage.<br/>**It's only used when the storage type is `Gcs`**. |
+| `storage.credential` | String | Unset | The credential of the google cloud storage.<br/>**It's only used when the storage type is `Gcs`**. |
+| `storage.container` | String | Unset | The container of the azure account.<br/>**It's only used when the storage type is `Azblob`**. |
+| `storage.sas_token` | String | Unset | The sas token of the azure account.<br/>**It's only used when the storage type is `Azblob`**. |
+| `storage.endpoint` | String | Unset | The endpoint of the S3 service.<br/>**It's only used when the storage type is `S3`, `Oss`, `Gcs` and `Azblob`**. |
+| `storage.region` | String | Unset | The region of the S3 service.<br/>**It's only used when the storage type is `S3`, `Oss`, `Gcs` and `Azblob`**. |
+| `storage.http_client` | -- | -- | The http client options to the storage.<br/>**It's only used when the storage type is `S3`, `Oss`, `Gcs` and `Azblob`**. |
+| `storage.http_client.pool_max_idle_per_host` | Integer | `1024` | The maximum idle connection per host allowed in the pool. |
+| `storage.http_client.connect_timeout` | String | `30s` | The timeout for only the connect phase of a http client. |
+| `storage.http_client.timeout` | String | `30s` | The total request timeout, applied from when the request starts connecting until the response body has finished.<br/>Also considered a total deadline. |
+| `storage.http_client.pool_idle_timeout` | String | `90s` | The timeout for idle sockets being kept-alive. |
 | `[[region_engine]]` | -- | -- | The region engine options. You can configure multiple region engines. |
 | `region_engine.mito` | -- | -- | The Mito engine options. |
 | `region_engine.mito.num_workers` | Integer | `8` | Number of region workers. |
@@ -393,21 +448,24 @@
 | `region_engine.mito.worker_request_batch_size` | Integer | `64` | Max batch size for a worker to handle requests. |
 | `region_engine.mito.manifest_checkpoint_distance` | Integer | `10` | Number of meta action updated to trigger a new checkpoint for the manifest. |
 | `region_engine.mito.compress_manifest` | Bool | `false` | Whether to compress manifest and checkpoint file by gzip (default false). |
-| `region_engine.mito.max_background_jobs` | Integer | `4` | Max number of running background jobs |
+| `region_engine.mito.max_background_flushes` | Integer | Auto | Max number of running background flush jobs (default: 1/2 of cpu cores). |
+| `region_engine.mito.max_background_compactions` | Integer | Auto | Max number of running background compaction jobs (default: 1/4 of cpu cores). |
+| `region_engine.mito.max_background_purges` | Integer | Auto | Max number of running background purge jobs (default: number of cpu cores). |
 | `region_engine.mito.auto_flush_interval` | String | `1h` | Interval to auto flush a region if it has not flushed yet. |
-| `region_engine.mito.global_write_buffer_size` | String | `1GB` | Global write buffer size for all regions. If not set, it's default to 1/8 of OS memory with a max limitation of 1GB. |
-| `region_engine.mito.global_write_buffer_reject_size` | String | `2GB` | Global write buffer size threshold to reject write requests. If not set, it's default to 2 times of `global_write_buffer_size` |
-| `region_engine.mito.sst_meta_cache_size` | String | `128MB` | Cache size for SST metadata. Setting it to 0 to disable the cache.<br/>If not set, it's default to 1/32 of OS memory with a max limitation of 128MB. |
-| `region_engine.mito.vector_cache_size` | String | `512MB` | Cache size for vectors and arrow arrays. Setting it to 0 to disable the cache.<br/>If not set, it's default to 1/16 of OS memory with a max limitation of 512MB. |
-| `region_engine.mito.page_cache_size` | String | `512MB` | Cache size for pages of SST row groups. Setting it to 0 to disable the cache.<br/>If not set, it's default to 1/16 of OS memory with a max limitation of 512MB. |
-| `region_engine.mito.enable_experimental_write_cache` | Bool | `false` | Whether to enable the experimental write cache. |
-| `region_engine.mito.experimental_write_cache_path` | String | `""` | File system path for write cache, defaults to `{data_home}/write_cache`. |
-| `region_engine.mito.experimental_write_cache_size` | String | `512MB` | Capacity for write cache. |
-| `region_engine.mito.experimental_write_cache_ttl` | String | `1h` | TTL for write cache. |
+| `region_engine.mito.global_write_buffer_size` | String | Auto | Global write buffer size for all regions. If not set, it's default to 1/8 of OS memory with a max limitation of 1GB. |
+| `region_engine.mito.global_write_buffer_reject_size` | String | Auto | Global write buffer size threshold to reject write requests. If not set, it's default to 2 times of `global_write_buffer_size` |
+| `region_engine.mito.sst_meta_cache_size` | String | Auto | Cache size for SST metadata. Setting it to 0 to disable the cache.<br/>If not set, it's default to 1/32 of OS memory with a max limitation of 128MB. |
+| `region_engine.mito.vector_cache_size` | String | Auto | Cache size for vectors and arrow arrays. Setting it to 0 to disable the cache.<br/>If not set, it's default to 1/16 of OS memory with a max limitation of 512MB. |
+| `region_engine.mito.page_cache_size` | String | Auto | Cache size for pages of SST row groups. Setting it to 0 to disable the cache.<br/>If not set, it's default to 1/8 of OS memory. |
+| `region_engine.mito.selector_result_cache_size` | String | Auto | Cache size for time series selector (e.g. `last_value()`). Setting it to 0 to disable the cache.<br/>If not set, it's default to 1/16 of OS memory with a max limitation of 512MB. |
+| `region_engine.mito.enable_experimental_write_cache` | Bool | `false` | Whether to enable the experimental write cache, it's enabled by default when using object storage. It is recommended to enable it when using object storage for better performance. |
+| `region_engine.mito.experimental_write_cache_path` | String | `""` | File system path for write cache, defaults to `{data_home}/object_cache/write`. |
+| `region_engine.mito.experimental_write_cache_size` | String | `5GiB` | Capacity for write cache. If your disk space is sufficient, it is recommended to set it larger. |
+| `region_engine.mito.experimental_write_cache_ttl` | String | Unset | TTL for write cache. |
 | `region_engine.mito.sst_write_buffer_size` | String | `8MB` | Buffer size for SST writing. |
-| `region_engine.mito.scan_parallelism` | Integer | `0` | Parallelism to scan a region (default: 1/4 of cpu cores).<br/>- `0`: using the default value (1/4 of cpu cores).<br/>- `1`: scan in current thread.<br/>- `n`: scan in parallelism n. |
 | `region_engine.mito.parallel_scan_channel_size` | Integer | `32` | Capacity of the channel to send data from parallel scan tasks to the main task. |
 | `region_engine.mito.allow_stale_entries` | Bool | `false` | Whether to allow stale WAL entries read during replay. |
+| `region_engine.mito.min_compaction_interval` | String | `0m` | Minimum time interval between two compactions.<br/>To align with the old behavior, the default value is 0 (no restrictions). |
 | `region_engine.mito.index` | -- | -- | The options for index in Mito engine. |
 | `region_engine.mito.index.aux_path` | String | `""` | Auxiliary directory path for the index in filesystem, used to store intermediate files for<br/>creating the index and staging files for searching the index, defaults to `{data_home}/index_intermediate`.<br/>The default name for this directory is `index_intermediate` for backward compatibility.<br/><br/>This path contains two subdirectories:<br/>- `__intm`: for storing intermediate files used during creating index.<br/>- `staging`: for storing staging files used during searching index. |
 | `region_engine.mito.index.staging_size` | String | `2GB` | The max capacity of the staging directory. |
@@ -429,23 +487,29 @@
 | `region_engine.mito.memtable.fork_dictionary_bytes` | String | `1GiB` | Max dictionary bytes.<br/>Only available for `partition_tree` memtable. |
 | `region_engine.file` | -- | -- | Enable the file engine. |
 | `logging` | -- | -- | The logging options. |
-| `logging.dir` | String | `/tmp/greptimedb/logs` | The directory to store the log files. |
-| `logging.level` | String | `None` | The log level. Can be `info`/`debug`/`warn`/`error`. |
+| `logging.dir` | String | `/tmp/greptimedb/logs` | The directory to store the log files. If set to empty, logs will not be written to files. |
+| `logging.level` | String | Unset | The log level. Can be `info`/`debug`/`warn`/`error`. |
 | `logging.enable_otlp_tracing` | Bool | `false` | Enable OTLP tracing. |
 | `logging.otlp_endpoint` | String | `http://localhost:4317` | The OTLP tracing endpoint. |
 | `logging.append_stdout` | Bool | `true` | Whether to append logs to stdout. |
+| `logging.log_format` | String | `text` | The log format. Can be `text`/`json`. |
+| `logging.max_log_files` | Integer | `720` | The maximum amount of log files. |
 | `logging.tracing_sample_ratio` | -- | -- | The percentage of tracing will be sampled and exported.<br/>Valid range `[0, 1]`, 1 means all traces are sampled, 0 means all traces are not sampled, the default value is 1.<br/>ratio > 1 are treated as 1. Fractions < 0 are treated as 0 |
 | `logging.tracing_sample_ratio.default_ratio` | Float | `1.0` | -- |
+| `logging.slow_query` | -- | -- | The slow query log options. |
+| `logging.slow_query.enable` | Bool | `false` | Whether to enable slow query log. |
+| `logging.slow_query.threshold` | String | Unset | The threshold of slow query. |
+| `logging.slow_query.sample_ratio` | Float | Unset | The sampling ratio of slow query log. The value should be in the range of (0, 1]. |
 | `export_metrics` | -- | -- | The datanode can export its metrics and send to Prometheus compatible service (e.g. send to `greptimedb` itself) from remote-write API.<br/>This is only used for `greptimedb` to export its own metrics internally. It's different from prometheus scrape. |
 | `export_metrics.enable` | Bool | `false` | whether enable export metrics. |
 | `export_metrics.write_interval` | String | `30s` | The interval of export metrics. |
-| `export_metrics.self_import` | -- | -- | For `standalone` mode, `self_import` is recommend to collect metrics generated by itself |
-| `export_metrics.self_import.db` | String | `None` | -- |
+| `export_metrics.self_import` | -- | -- | For `standalone` mode, `self_import` is recommend to collect metrics generated by itself<br/>You must create the database before enabling it. |
+| `export_metrics.self_import.db` | String | Unset | -- |
 | `export_metrics.remote_write` | -- | -- | -- |
-| `export_metrics.remote_write.url` | String | `""` | The url the metrics send to. The url example can be: `http://127.0.0.1:4000/v1/prometheus/write?db=information_schema`. |
+| `export_metrics.remote_write.url` | String | `""` | The url the metrics send to. The url example can be: `http://127.0.0.1:4000/v1/prometheus/write?db=greptime_metrics`. |
 | `export_metrics.remote_write.headers` | InlineTable | -- | HTTP headers of Prometheus remote-write carry. |
 | `tracing` | -- | -- | The tracing options. Only effect when compiled with `tokio-console` feature. |
-| `tracing.tokio_console_addr` | String | `None` | The tokio console address. |
+| `tracing.tokio_console_addr` | String | Unset | The tokio console address. |
 
 
 ### Flownode
@@ -453,7 +517,7 @@
 | Key | Type | Default | Descriptions |
 | --- | -----| ------- | ----------- |
 | `mode` | String | `distributed` | The running mode of the flownode. It can be `standalone` or `distributed`. |
-| `node_id` | Integer | `None` | The flownode identifier and should be unique in the cluster. |
+| `node_id` | Integer | Unset | The flownode identifier and should be unique in the cluster. |
 | `grpc` | -- | -- | The gRPC server options. |
 | `grpc.addr` | String | `127.0.0.1:6800` | The address to bind the gRPC server. |
 | `grpc.hostname` | String | `127.0.0.1` | The hostname advertised to the metasrv,<br/>and used for connections from outside the host |
@@ -474,12 +538,18 @@
 | `heartbeat.interval` | String | `3s` | Interval for sending heartbeat messages to the metasrv. |
 | `heartbeat.retry_interval` | String | `3s` | Interval for retrying to send heartbeat messages to the metasrv. |
 | `logging` | -- | -- | The logging options. |
-| `logging.dir` | String | `/tmp/greptimedb/logs` | The directory to store the log files. |
-| `logging.level` | String | `None` | The log level. Can be `info`/`debug`/`warn`/`error`. |
+| `logging.dir` | String | `/tmp/greptimedb/logs` | The directory to store the log files. If set to empty, logs will not be written to files. |
+| `logging.level` | String | Unset | The log level. Can be `info`/`debug`/`warn`/`error`. |
 | `logging.enable_otlp_tracing` | Bool | `false` | Enable OTLP tracing. |
 | `logging.otlp_endpoint` | String | `http://localhost:4317` | The OTLP tracing endpoint. |
 | `logging.append_stdout` | Bool | `true` | Whether to append logs to stdout. |
+| `logging.log_format` | String | `text` | The log format. Can be `text`/`json`. |
+| `logging.max_log_files` | Integer | `720` | The maximum amount of log files. |
 | `logging.tracing_sample_ratio` | -- | -- | The percentage of tracing will be sampled and exported.<br/>Valid range `[0, 1]`, 1 means all traces are sampled, 0 means all traces are not sampled, the default value is 1.<br/>ratio > 1 are treated as 1. Fractions < 0 are treated as 0 |
 | `logging.tracing_sample_ratio.default_ratio` | Float | `1.0` | -- |
+| `logging.slow_query` | -- | -- | The slow query log options. |
+| `logging.slow_query.enable` | Bool | `false` | Whether to enable slow query log. |
+| `logging.slow_query.threshold` | String | Unset | The threshold of slow query. |
+| `logging.slow_query.sample_ratio` | Float | Unset | The sampling ratio of slow query log. The value should be in the range of (0, 1]. |
 | `tracing` | -- | -- | The tracing options. Only effect when compiled with `tokio-console` feature. |
-| `tracing.tokio_console_addr` | String | `None` | The tokio console address. |
+| `tracing.tokio_console_addr` | String | Unset | The tokio console address. |

config/datanode.example.toml

@@ -2,7 +2,7 @@
 mode = "standalone"
 
 ## The datanode identifier and should be unique in the cluster.
-## +toml2docs:none-default
+## @toml2docs:none-default
 node_id = 42
 
 ## Start services after regions have obtained leases.
@@ -19,26 +19,41 @@ enable_telemetry = true
 ## Parallelism of initializing regions.
 init_regions_parallelism = 16
 
+## The maximum current queries allowed to be executed. Zero means unlimited.
+max_concurrent_queries = 0
+
 ## Deprecated, use `grpc.addr` instead.
-## +toml2docs:none-default
+## @toml2docs:none-default
 rpc_addr = "127.0.0.1:3001"
 
 ## Deprecated, use `grpc.hostname` instead.
-## +toml2docs:none-default
+## @toml2docs:none-default
 rpc_hostname = "127.0.0.1"
 
 ## Deprecated, use `grpc.runtime_size` instead.
-## +toml2docs:none-default
+## @toml2docs:none-default
 rpc_runtime_size = 8
 
 ## Deprecated, use `grpc.rpc_max_recv_message_size` instead.
-## +toml2docs:none-default
+## @toml2docs:none-default
 rpc_max_recv_message_size = "512MB"
 
 ## Deprecated, use `grpc.rpc_max_send_message_size` instead.
-## +toml2docs:none-default
+## @toml2docs:none-default
 rpc_max_send_message_size = "512MB"
 
+
+## The HTTP server options.
+[http]
+## The address to bind the HTTP server.
+addr = "127.0.0.1:4000"
+## HTTP request timeout. Set to 0 to disable timeout.
+timeout = "30s"
+## HTTP request body limit.
+## The following units are supported: `B`, `KB`, `KiB`, `MB`, `MiB`, `GB`, `GiB`, `TB`, `TiB`, `PB`, `PiB`.
+## Set to 0 to disable limit.
+body_limit = "64MB"
+
 ## The gRPC server options.
 [grpc]
 ## The address to bind the gRPC server.
@@ -59,11 +74,11 @@ max_send_message_size = "512MB"
 mode = "disable"
 
 ## Certificate file path.
-## +toml2docs:none-default
+## @toml2docs:none-default
 cert_path = ""
 
 ## Private key file path.
-## +toml2docs:none-default
+## @toml2docs:none-default
 key_path = ""
 
 ## Watch for Certificate and key file change and auto reload.
@@ -71,11 +86,11 @@ key_path = ""
 watch = false
 
 ## The runtime options.
-[runtime]
+#+ [runtime]
 ## The number of threads to execute the runtime for global read operations.
-global_rt_size = 8
+#+ global_rt_size = 8
 ## The number of threads to execute the runtime for global write operations.
-compact_rt_size = 4
+#+ compact_rt_size = 4
 
 ## The heartbeat options.
 [heartbeat]
@@ -123,7 +138,7 @@ provider = "raft_engine"
 
 ## The directory to store the WAL files.
 ## **It's only used when the provider is `raft_engine`**.
-## +toml2docs:none-default
+## @toml2docs:none-default
 dir = "/tmp/greptimedb/wal"
 
 ## The size of the WAL segment file.
@@ -158,6 +173,9 @@ prefill_log_files = false
 ## **It's only used when the provider is `raft_engine`**.
 sync_period = "10s"
 
+## Parallelism during WAL recovery.
+recovery_parallelism = 2
+
 ## The Kafka broker endpoints.
 ## **It's only used when the provider is `kafka`**.
 broker_endpoints = ["127.0.0.1:9092"]
@@ -187,6 +205,43 @@ backoff_base = 2
 ## **It's only used when the provider is `kafka`**.
 backoff_deadline = "5mins"
 
+## Whether to enable WAL index creation.
+## **It's only used when the provider is `kafka`**.
+create_index = true
+
+## The interval for dumping WAL indexes.
+## **It's only used when the provider is `kafka`**.
+dump_index_interval = "60s"
+
+## Ignore missing entries during read WAL.
+## **It's only used when the provider is `kafka`**.
+##
+## This option ensures that when Kafka messages are deleted, the system
+## can still successfully replay memtable data without throwing an
+## out-of-range error.
+## However, enabling this option might lead to unexpected data loss,
+## as the system will skip over missing entries instead of treating
+## them as critical errors.
+overwrite_entry_start_id = false
+
+# The Kafka SASL configuration.
+# **It's only used when the provider is `kafka`**.
+# Available SASL mechanisms:
+# - `PLAIN`
+# - `SCRAM-SHA-256`
+# - `SCRAM-SHA-512`
+# [wal.sasl]
+# type = "SCRAM-SHA-512"
+# username = "user_kafka"
+# password = "secret"
+
+# The Kafka TLS configuration.
+# **It's only used when the provider is `kafka`**.
+# [wal.tls]
+# server_ca_cert_path = "/path/to/server_cert"
+# client_cert_path = "/path/to/client_cert"
+# client_key_path = "/path/to/key"
+
 # Example of using S3 as the storage.
 # [storage]
 # type = "S3"
@@ -223,6 +278,7 @@ backoff_deadline = "5mins"
 # root = "data"
 # scope = "test"
 # credential_path = "123456"
+# credential = "base64-credential"
 # endpoint = "https://storage.googleapis.com"
 
 ## The data storage options.
@@ -238,87 +294,123 @@ data_home = "/tmp/greptimedb/"
 ## - `Oss`: the data is stored in the Aliyun OSS.
 type = "File"
 
-## Cache configuration for object storage such as 'S3' etc.
-## The local file cache directory.
-## +toml2docs:none-default
-cache_path = "/path/local_cache"
+## Read cache configuration for object storage such as 'S3' etc, it's configured by default when using object storage. It is recommended to configure it when using object storage for better performance.
+## A local file directory, defaults to `{data_home}/object_cache/read`. An empty string means disabling.
+## @toml2docs:none-default
+#+ cache_path = ""
 
-## The local file cache capacity in bytes.
-## +toml2docs:none-default
-cache_capacity = "256MB"
+## The local file cache capacity in bytes. If your disk space is sufficient, it is recommended to set it larger.
+## @toml2docs:none-default
+cache_capacity = "5GiB"
 
 ## The S3 bucket name.
 ## **It's only used when the storage type is `S3`, `Oss` and `Gcs`**.
-## +toml2docs:none-default
+## @toml2docs:none-default
 bucket = "greptimedb"
 
 ## The S3 data will be stored in the specified prefix, for example, `s3://${bucket}/${root}`.
 ## **It's only used when the storage type is `S3`, `Oss` and `Azblob`**.
-## +toml2docs:none-default
+## @toml2docs:none-default
 root = "greptimedb"
 
 ## The access key id of the aws account.
 ## It's **highly recommended** to use AWS IAM roles instead of hardcoding the access key id and secret key.
 ## **It's only used when the storage type is `S3` and `Oss`**.
-## +toml2docs:none-default
+## @toml2docs:none-default
 access_key_id = "test"
 
 ## The secret access key of the aws account.
 ## It's **highly recommended** to use AWS IAM roles instead of hardcoding the access key id and secret key.
 ## **It's only used when the storage type is `S3`**.
-## +toml2docs:none-default
+## @toml2docs:none-default
 secret_access_key = "test"
 
 ## The secret access key of the aliyun account.
 ## **It's only used when the storage type is `Oss`**.
-## +toml2docs:none-default
+## @toml2docs:none-default
 access_key_secret = "test"
 
 ## The account key of the azure account.
 ## **It's only used when the storage type is `Azblob`**.
-## +toml2docs:none-default
+## @toml2docs:none-default
 account_name = "test"
 
 ## The account key of the azure account.
 ## **It's only used when the storage type is `Azblob`**.
-## +toml2docs:none-default
+## @toml2docs:none-default
 account_key = "test"
 
 ## The scope of the google cloud storage.
 ## **It's only used when the storage type is `Gcs`**.
-## +toml2docs:none-default
+## @toml2docs:none-default
 scope = "test"
 
 ## The credential path of the google cloud storage.
 ## **It's only used when the storage type is `Gcs`**.
-## +toml2docs:none-default
+## @toml2docs:none-default
 credential_path = "test"
 
+## The credential of the google cloud storage.
+## **It's only used when the storage type is `Gcs`**.
+## @toml2docs:none-default
+credential = "base64-credential"
+
 ## The container of the azure account.
 ## **It's only used when the storage type is `Azblob`**.
-## +toml2docs:none-default
+## @toml2docs:none-default
 container = "greptimedb"
 
 ## The sas token of the azure account.
 ## **It's only used when the storage type is `Azblob`**.
-## +toml2docs:none-default
+## @toml2docs:none-default
 sas_token = ""
 
 ## The endpoint of the S3 service.
 ## **It's only used when the storage type is `S3`, `Oss`, `Gcs` and `Azblob`**.
-## +toml2docs:none-default
+## @toml2docs:none-default
 endpoint = "https://s3.amazonaws.com"
 
 ## The region of the S3 service.
 ## **It's only used when the storage type is `S3`, `Oss`, `Gcs` and `Azblob`**.
-## +toml2docs:none-default
+## @toml2docs:none-default
 region = "us-west-2"
 
+## The http client options to the storage.
+## **It's only used when the storage type is `S3`, `Oss`, `Gcs` and `Azblob`**.
+[storage.http_client]
+
+## The maximum idle connection per host allowed in the pool.
+pool_max_idle_per_host = 1024
+
+## The timeout for only the connect phase of a http client.
+connect_timeout = "30s"
+
+## The total request timeout, applied from when the request starts connecting until the response body has finished.
+## Also considered a total deadline.
+timeout = "30s"
+
+## The timeout for idle sockets being kept-alive.
+pool_idle_timeout = "90s"
+
 # Custom storage options
 # [[storage.providers]]
+# name = "S3"
 # type = "S3"
+# bucket = "greptimedb"
+# root = "data"
+# access_key_id = "test"
+# secret_access_key = "123456"
+# endpoint = "https://s3.amazonaws.com"
+# region = "us-west-2"
 # [[storage.providers]]
+# name = "Gcs"
 # type = "Gcs"
+# bucket = "greptimedb"
+# root = "data"
+# scope = "test"
+# credential_path = "123456"
+# credential = "base64-credential"
+# endpoint = "https://storage.googleapis.com"
 
 ## The region engine options. You can configure multiple region engines.
 [[region_engine]]
@@ -327,7 +419,7 @@ region = "us-west-2"
 [region_engine.mito]
 
 ## Number of region workers.
-num_workers = 8
+#+ num_workers = 8
 
 ## Request channel size of each worker.
 worker_channel_size = 128
@@ -341,57 +433,75 @@ manifest_checkpoint_distance = 10
 ## Whether to compress manifest and checkpoint file by gzip (default false).
 compress_manifest = false
 
-## Max number of running background jobs
-max_background_jobs = 4
+## Max number of running background flush jobs (default: 1/2 of cpu cores).
+## @toml2docs:none-default="Auto"
+#+ max_background_flushes = 4
+
+## Max number of running background compaction jobs (default: 1/4 of cpu cores).
+## @toml2docs:none-default="Auto"
+#+ max_background_compactions = 2
+
+## Max number of running background purge jobs (default: number of cpu cores).
+## @toml2docs:none-default="Auto"
+#+ max_background_purges = 8
 
 ## Interval to auto flush a region if it has not flushed yet.
 auto_flush_interval = "1h"
 
 ## Global write buffer size for all regions. If not set, it's default to 1/8 of OS memory with a max limitation of 1GB.
-global_write_buffer_size = "1GB"
+## @toml2docs:none-default="Auto"
+#+ global_write_buffer_size = "1GB"
 
 ## Global write buffer size threshold to reject write requests. If not set, it's default to 2 times of `global_write_buffer_size`
-global_write_buffer_reject_size = "2GB"
+## @toml2docs:none-default="Auto"
+#+ global_write_buffer_reject_size = "2GB"
 
 ## Cache size for SST metadata. Setting it to 0 to disable the cache.
 ## If not set, it's default to 1/32 of OS memory with a max limitation of 128MB.
-sst_meta_cache_size = "128MB"
+## @toml2docs:none-default="Auto"
+#+ sst_meta_cache_size = "128MB"
 
 ## Cache size for vectors and arrow arrays. Setting it to 0 to disable the cache.
 ## If not set, it's default to 1/16 of OS memory with a max limitation of 512MB.
-vector_cache_size = "512MB"
+## @toml2docs:none-default="Auto"
+#+ vector_cache_size = "512MB"
 
 ## Cache size for pages of SST row groups. Setting it to 0 to disable the cache.
-## If not set, it's default to 1/16 of OS memory with a max limitation of 512MB.
-page_cache_size = "512MB"
+## If not set, it's default to 1/8 of OS memory.
+## @toml2docs:none-default="Auto"
+#+ page_cache_size = "512MB"
 
-## Whether to enable the experimental write cache.
+## Cache size for time series selector (e.g. `last_value()`). Setting it to 0 to disable the cache.
+## If not set, it's default to 1/16 of OS memory with a max limitation of 512MB.
+## @toml2docs:none-default="Auto"
+#+ selector_result_cache_size = "512MB"
+
+## Whether to enable the experimental write cache, it's enabled by default when using object storage. It is recommended to enable it when using object storage for better performance.
 enable_experimental_write_cache = false
 
-## File system path for write cache, defaults to `{data_home}/write_cache`.
+## File system path for write cache, defaults to `{data_home}/object_cache/write`.
 experimental_write_cache_path = ""
 
-## Capacity for write cache.
-experimental_write_cache_size = "512MB"
+## Capacity for write cache. If your disk space is sufficient, it is recommended to set it larger.
+experimental_write_cache_size = "5GiB"
 
 ## TTL for write cache.
-experimental_write_cache_ttl = "1h"
+## @toml2docs:none-default
+experimental_write_cache_ttl = "8h"
 
 ## Buffer size for SST writing.
 sst_write_buffer_size = "8MB"
 
-## Parallelism to scan a region (default: 1/4 of cpu cores).
-## - `0`: using the default value (1/4 of cpu cores).
-## - `1`: scan in current thread.
-## - `n`: scan in parallelism n.
-scan_parallelism = 0
-
 ## Capacity of the channel to send data from parallel scan tasks to the main task.
 parallel_scan_channel_size = 32
 
 ## Whether to allow stale WAL entries read during replay.
 allow_stale_entries = false
 
+## Minimum time interval between two compactions.
+## To align with the old behavior, the default value is 0 (no restrictions).
+min_compaction_interval = "0m"
+
 ## The options for index in Mito engine.
 [region_engine.mito.index]
 
@@ -482,11 +592,11 @@ fork_dictionary_bytes = "1GiB"
 
 ## The logging options.
 [logging]
-## The directory to store the log files.
+## The directory to store the log files. If set to empty, logs will not be written to files.
 dir = "/tmp/greptimedb/logs"
 
 ## The log level. Can be `info`/`debug`/`warn`/`error`.
-## +toml2docs:none-default
+## @toml2docs:none-default
 level = "info"
 
 ## Enable OTLP tracing.
@@ -498,12 +608,31 @@ otlp_endpoint = "http://localhost:4317"
 ## Whether to append logs to stdout.
 append_stdout = true
 
+## The log format. Can be `text`/`json`.
+log_format = "text"
+
+## The maximum amount of log files.
+max_log_files = 720
+
 ## The percentage of tracing will be sampled and exported.
 ## Valid range `[0, 1]`, 1 means all traces are sampled, 0 means all traces are not sampled, the default value is 1.
 ## ratio > 1 are treated as 1. Fractions < 0 are treated as 0
 [logging.tracing_sample_ratio]
 default_ratio = 1.0
 
+## The slow query log options.
+[logging.slow_query]
+## Whether to enable slow query log.
+enable = false
+
+## The threshold of slow query.
+## @toml2docs:none-default
+threshold = "10s"
+
+## The sampling ratio of slow query log. The value should be in the range of (0, 1].
+## @toml2docs:none-default
+sample_ratio = 1.0
+
 ## The datanode can export its metrics and send to Prometheus compatible service (e.g. send to `greptimedb` itself) from remote-write API.
 ## This is only used for `greptimedb` to export its own metrics internally. It's different from prometheus scrape.
 [export_metrics]
@@ -515,19 +644,20 @@ enable = false
 write_interval = "30s"
 
 ## For `standalone` mode, `self_import` is recommend to collect metrics generated by itself
+## You must create the database before enabling it.
 [export_metrics.self_import]
-## +toml2docs:none-default
-db = "information_schema"
+## @toml2docs:none-default
+db = "greptime_metrics"
 
 [export_metrics.remote_write]
-## The url the metrics send to. The url example can be: `http://127.0.0.1:4000/v1/prometheus/write?db=information_schema`.
+## The url the metrics send to. The url example can be: `http://127.0.0.1:4000/v1/prometheus/write?db=greptime_metrics`.
 url = ""
 
 ## HTTP headers of Prometheus remote-write carry.
 headers = { }
 
 ## The tracing options. Only effect when compiled with `tokio-console` feature.
-[tracing]
+#+ [tracing]
 ## The tokio console address.
-## +toml2docs:none-default
-tokio_console_addr = "127.0.0.1"
+## @toml2docs:none-default
+#+ tokio_console_addr = "127.0.0.1"

config/flownode.example.toml

@@ -2,7 +2,7 @@
 mode = "distributed"
 
 ## The flownode identifier and should be unique in the cluster.
-## +toml2docs:none-default
+## @toml2docs:none-default
 node_id = 14
 
 ## The gRPC server options.
@@ -59,11 +59,11 @@ retry_interval = "3s"
 
 ## The logging options.
 [logging]
-## The directory to store the log files.
+## The directory to store the log files. If set to empty, logs will not be written to files.
 dir = "/tmp/greptimedb/logs"
 
 ## The log level. Can be `info`/`debug`/`warn`/`error`.
-## +toml2docs:none-default
+## @toml2docs:none-default
 level = "info"
 
 ## Enable OTLP tracing.
@@ -75,15 +75,34 @@ otlp_endpoint = "http://localhost:4317"
 ## Whether to append logs to stdout.
 append_stdout = true
 
+## The log format. Can be `text`/`json`.
+log_format = "text"
+
+## The maximum amount of log files.
+max_log_files = 720
+
 ## The percentage of tracing will be sampled and exported.
 ## Valid range `[0, 1]`, 1 means all traces are sampled, 0 means all traces are not sampled, the default value is 1.
 ## ratio > 1 are treated as 1. Fractions < 0 are treated as 0
 [logging.tracing_sample_ratio]
 default_ratio = 1.0
 
-## The tracing options. Only effect when compiled with `tokio-console` feature.
-[tracing]
-## The tokio console address.
-## +toml2docs:none-default
-tokio_console_addr = "127.0.0.1"
+## The slow query log options.
+[logging.slow_query]
+## Whether to enable slow query log.
+enable = false
+
+## The threshold of slow query.
+## @toml2docs:none-default
+threshold = "10s"
+
+## The sampling ratio of slow query log. The value should be in the range of (0, 1].
+## @toml2docs:none-default
+sample_ratio = 1.0
+
+## The tracing options. Only effect when compiled with `tokio-console` feature.
+#+ [tracing]
+## The tokio console address.
+## @toml2docs:none-default
+#+ tokio_console_addr = "127.0.0.1"
 

config/frontend.example.toml

@@ -1,13 +1,13 @@
 ## The default timezone of the server.
-## +toml2docs:none-default
+## @toml2docs:none-default
 default_timezone = "UTC"
 
 ## The runtime options.
-[runtime]
+#+ [runtime]
 ## The number of threads to execute the runtime for global read operations.
-global_rt_size = 8
+#+ global_rt_size = 8
 ## The number of threads to execute the runtime for global write operations.
-compact_rt_size = 4
+#+ compact_rt_size = 4
 
 ## The heartbeat options.
 [heartbeat]
@@ -44,11 +44,11 @@ runtime_size = 8
 mode = "disable"
 
 ## Certificate file path.
-## +toml2docs:none-default
+## @toml2docs:none-default
 cert_path = ""
 
 ## Private key file path.
-## +toml2docs:none-default
+## @toml2docs:none-default
 key_path = ""
 
 ## Watch for Certificate and key file change and auto reload.
@@ -76,11 +76,11 @@ runtime_size = 2
 mode = "disable"
 
 ## Certificate file path.
-## +toml2docs:none-default
+## @toml2docs:none-default
 cert_path = ""
 
 ## Private key file path.
-## +toml2docs:none-default
+## @toml2docs:none-default
 key_path = ""
 
 ## Watch for Certificate and key file change and auto reload
@@ -101,11 +101,11 @@ runtime_size = 2
 mode = "disable"
 
 ## Certificate file path.
-## +toml2docs:none-default
+## @toml2docs:none-default
 cert_path = ""
 
 ## Private key file path.
-## +toml2docs:none-default
+## @toml2docs:none-default
 key_path = ""
 
 ## Watch for Certificate and key file change and auto reload
@@ -166,11 +166,11 @@ tcp_nodelay = true
 
 ## The logging options.
 [logging]
-## The directory to store the log files.
+## The directory to store the log files. If set to empty, logs will not be written to files.
 dir = "/tmp/greptimedb/logs"
 
 ## The log level. Can be `info`/`debug`/`warn`/`error`.
-## +toml2docs:none-default
+## @toml2docs:none-default
 level = "info"
 
 ## Enable OTLP tracing.
@@ -182,12 +182,31 @@ otlp_endpoint = "http://localhost:4317"
 ## Whether to append logs to stdout.
 append_stdout = true
 
+## The log format. Can be `text`/`json`.
+log_format = "text"
+
+## The maximum amount of log files.
+max_log_files = 720
+
 ## The percentage of tracing will be sampled and exported.
 ## Valid range `[0, 1]`, 1 means all traces are sampled, 0 means all traces are not sampled, the default value is 1.
 ## ratio > 1 are treated as 1. Fractions < 0 are treated as 0
 [logging.tracing_sample_ratio]
 default_ratio = 1.0
 
+## The slow query log options.
+[logging.slow_query]
+## Whether to enable slow query log.
+enable = false
+
+## The threshold of slow query.
+## @toml2docs:none-default
+threshold = "10s"
+
+## The sampling ratio of slow query log. The value should be in the range of (0, 1].
+## @toml2docs:none-default
+sample_ratio = 1.0
+
 ## The datanode can export its metrics and send to Prometheus compatible service (e.g. send to `greptimedb` itself) from remote-write API.
 ## This is only used for `greptimedb` to export its own metrics internally. It's different from prometheus scrape.
 [export_metrics]
@@ -199,19 +218,20 @@ enable = false
 write_interval = "30s"
 
 ## For `standalone` mode, `self_import` is recommend to collect metrics generated by itself
+## You must create the database before enabling it.
 [export_metrics.self_import]
-## +toml2docs:none-default
-db = "information_schema"
+## @toml2docs:none-default
+db = "greptime_metrics"
 
 [export_metrics.remote_write]
-## The url the metrics send to. The url example can be: `http://127.0.0.1:4000/v1/prometheus/write?db=information_schema`.
+## The url the metrics send to. The url example can be: `http://127.0.0.1:4000/v1/prometheus/write?db=greptime_metrics`.
 url = ""
 
 ## HTTP headers of Prometheus remote-write carry.
 headers = { }
 
 ## The tracing options. Only effect when compiled with `tokio-console` feature.
-[tracing]
+#+ [tracing]
 ## The tokio console address.
-## +toml2docs:none-default
-tokio_console_addr = "127.0.0.1"
+## @toml2docs:none-default
+#+ tokio_console_addr = "127.0.0.1"

config/metasrv.example.toml

@@ -7,8 +7,8 @@ bind_addr = "127.0.0.1:3002"
 ## The communication server address for frontend and datanode to connect to metasrv,  "127.0.0.1:3002" by default for localhost.
 server_addr = "127.0.0.1:3002"
 
-## Etcd server address.
-store_addr = "127.0.0.1:2379"
+## Store server address default to etcd store.
+store_addrs = ["127.0.0.1:2379"]
 
 ## Datanode selector type.
 ## - `round_robin` (default value)
@@ -32,12 +32,15 @@ store_key_prefix = ""
 ## - Using shared storage (e.g., s3).
 enable_region_failover = false
 
+## The datastore for meta server.
+backend = "EtcdStore"
+
 ## The runtime options.
-[runtime]
+#+ [runtime]
 ## The number of threads to execute the runtime for global read operations.
-global_rt_size = 8
+#+ global_rt_size = 8
 ## The number of threads to execute the runtime for global write operations.
-compact_rt_size = 4
+#+ compact_rt_size = 4
 
 ## Procedure storage options.
 [procedure]
@@ -96,7 +99,12 @@ provider = "raft_engine"
 ## The broker endpoints of the Kafka cluster.
 broker_endpoints = ["127.0.0.1:9092"]
 
-## Number of topics to be created upon start.
+## Automatically create topics for WAL.
+## Set to `true` to automatically create topics for WAL.
+## Otherwise, use topics named `topic_name_prefix_[0..num_topics)`
+auto_create_topics = true
+
+## Number of topics.
 num_topics = 64
 
 ## Topic selector type.
@@ -105,6 +113,7 @@ num_topics = 64
 selector_type = "round_robin"
 
 ## A Kafka topic is constructed by concatenating `topic_name_prefix` and `topic_id`.
+## i.g., greptimedb_wal_topic_0, greptimedb_wal_topic_1.
 topic_name_prefix = "greptimedb_wal_topic"
 
 ## Expected number of replicas of each partition.
@@ -124,13 +133,31 @@ backoff_base = 2
 ## Stop reconnecting if the total wait time reaches the deadline. If this config is missing, the reconnecting won't terminate.
 backoff_deadline = "5mins"
 
+# The Kafka SASL configuration.
+# **It's only used when the provider is `kafka`**.
+# Available SASL mechanisms:
+# - `PLAIN`
+# - `SCRAM-SHA-256`
+# - `SCRAM-SHA-512`
+# [wal.sasl]
+# type = "SCRAM-SHA-512"
+# username = "user_kafka"
+# password = "secret"
+
+# The Kafka TLS configuration.
+# **It's only used when the provider is `kafka`**.
+# [wal.tls]
+# server_ca_cert_path = "/path/to/server_cert"
+# client_cert_path = "/path/to/client_cert"
+# client_key_path = "/path/to/key"
+
 ## The logging options.
 [logging]
-## The directory to store the log files.
+## The directory to store the log files. If set to empty, logs will not be written to files.
 dir = "/tmp/greptimedb/logs"
 
 ## The log level. Can be `info`/`debug`/`warn`/`error`.
-## +toml2docs:none-default
+## @toml2docs:none-default
 level = "info"
 
 ## Enable OTLP tracing.
@@ -142,12 +169,31 @@ otlp_endpoint = "http://localhost:4317"
 ## Whether to append logs to stdout.
 append_stdout = true
 
+## The log format. Can be `text`/`json`.
+log_format = "text"
+
+## The maximum amount of log files.
+max_log_files = 720
+
 ## The percentage of tracing will be sampled and exported.
 ## Valid range `[0, 1]`, 1 means all traces are sampled, 0 means all traces are not sampled, the default value is 1.
 ## ratio > 1 are treated as 1. Fractions < 0 are treated as 0
 [logging.tracing_sample_ratio]
 default_ratio = 1.0
 
+## The slow query log options.
+[logging.slow_query]
+## Whether to enable slow query log.
+enable = false
+
+## The threshold of slow query.
+## @toml2docs:none-default
+threshold = "10s"
+
+## The sampling ratio of slow query log. The value should be in the range of (0, 1].
+## @toml2docs:none-default
+sample_ratio = 1.0
+
 ## The datanode can export its metrics and send to Prometheus compatible service (e.g. send to `greptimedb` itself) from remote-write API.
 ## This is only used for `greptimedb` to export its own metrics internally. It's different from prometheus scrape.
 [export_metrics]
@@ -159,19 +205,20 @@ enable = false
 write_interval = "30s"
 
 ## For `standalone` mode, `self_import` is recommend to collect metrics generated by itself
+## You must create the database before enabling it.
 [export_metrics.self_import]
-## +toml2docs:none-default
-db = "information_schema"
+## @toml2docs:none-default
+db = "greptime_metrics"
 
 [export_metrics.remote_write]
-## The url the metrics send to. The url example can be: `http://127.0.0.1:4000/v1/prometheus/write?db=information_schema`.
+## The url the metrics send to. The url example can be: `http://127.0.0.1:4000/v1/prometheus/write?db=greptime_metrics`.
 url = ""
 
 ## HTTP headers of Prometheus remote-write carry.
 headers = { }
 
 ## The tracing options. Only effect when compiled with `tokio-console` feature.
-[tracing]
+#+ [tracing]
 ## The tokio console address.
-## +toml2docs:none-default
-tokio_console_addr = "127.0.0.1"
+## @toml2docs:none-default
+#+ tokio_console_addr = "127.0.0.1"

config/standalone.example.toml

@@ -5,15 +5,25 @@ mode = "standalone"
 enable_telemetry = true
 
 ## The default timezone of the server.
-## +toml2docs:none-default
+## @toml2docs:none-default
 default_timezone = "UTC"
 
+## Initialize all regions in the background during the startup.
+## By default, it provides services after all regions have been initialized.
+init_regions_in_background = false
+
+## Parallelism of initializing regions.
+init_regions_parallelism = 16
+
+## The maximum current queries allowed to be executed. Zero means unlimited.
+max_concurrent_queries = 0
+
 ## The runtime options.
-[runtime]
+#+ [runtime]
 ## The number of threads to execute the runtime for global read operations.
-global_rt_size = 8
+#+ global_rt_size = 8
 ## The number of threads to execute the runtime for global write operations.
-compact_rt_size = 4
+#+ compact_rt_size = 4
 
 ## The HTTP server options.
 [http]
@@ -39,11 +49,11 @@ runtime_size = 8
 mode = "disable"
 
 ## Certificate file path.
-## +toml2docs:none-default
+## @toml2docs:none-default
 cert_path = ""
 
 ## Private key file path.
-## +toml2docs:none-default
+## @toml2docs:none-default
 key_path = ""
 
 ## Watch for Certificate and key file change and auto reload.
@@ -71,11 +81,11 @@ runtime_size = 2
 mode = "disable"
 
 ## Certificate file path.
-## +toml2docs:none-default
+## @toml2docs:none-default
 cert_path = ""
 
 ## Private key file path.
-## +toml2docs:none-default
+## @toml2docs:none-default
 key_path = ""
 
 ## Watch for Certificate and key file change and auto reload
@@ -96,11 +106,11 @@ runtime_size = 2
 mode = "disable"
 
 ## Certificate file path.
-## +toml2docs:none-default
+## @toml2docs:none-default
 cert_path = ""
 
 ## Private key file path.
-## +toml2docs:none-default
+## @toml2docs:none-default
 key_path = ""
 
 ## Watch for Certificate and key file change and auto reload
@@ -132,7 +142,7 @@ provider = "raft_engine"
 
 ## The directory to store the WAL files.
 ## **It's only used when the provider is `raft_engine`**.
-## +toml2docs:none-default
+## @toml2docs:none-default
 dir = "/tmp/greptimedb/wal"
 
 ## The size of the WAL segment file.
@@ -167,10 +177,41 @@ prefill_log_files = false
 ## **It's only used when the provider is `raft_engine`**.
 sync_period = "10s"
 
+## Parallelism during WAL recovery.
+recovery_parallelism = 2
+
 ## The Kafka broker endpoints.
 ## **It's only used when the provider is `kafka`**.
 broker_endpoints = ["127.0.0.1:9092"]
 
+## Automatically create topics for WAL.
+## Set to `true` to automatically create topics for WAL.
+## Otherwise, use topics named `topic_name_prefix_[0..num_topics)`
+auto_create_topics = true
+
+## Number of topics.
+## **It's only used when the provider is `kafka`**.
+num_topics = 64
+
+## Topic selector type.
+## Available selector types:
+## - `round_robin` (default)
+## **It's only used when the provider is `kafka`**.
+selector_type = "round_robin"
+
+## A Kafka topic is constructed by concatenating `topic_name_prefix` and `topic_id`.
+## i.g., greptimedb_wal_topic_0, greptimedb_wal_topic_1.
+## **It's only used when the provider is `kafka`**.
+topic_name_prefix = "greptimedb_wal_topic"
+
+## Expected number of replicas of each partition.
+## **It's only used when the provider is `kafka`**.
+replication_factor = 1
+
+## Above which a topic creation operation will be cancelled.
+## **It's only used when the provider is `kafka`**.
+create_topic_timeout = "30s"
+
 ## The max size of a single producer batch.
 ## Warning: Kafka has a default limit of 1MB per message in a topic.
 ## **It's only used when the provider is `kafka`**.
@@ -196,6 +237,35 @@ backoff_base = 2
 ## **It's only used when the provider is `kafka`**.
 backoff_deadline = "5mins"
 
+## Ignore missing entries during read WAL.
+## **It's only used when the provider is `kafka`**.
+##
+## This option ensures that when Kafka messages are deleted, the system
+## can still successfully replay memtable data without throwing an
+## out-of-range error.
+## However, enabling this option might lead to unexpected data loss,
+## as the system will skip over missing entries instead of treating
+## them as critical errors.
+overwrite_entry_start_id = false
+
+# The Kafka SASL configuration.
+# **It's only used when the provider is `kafka`**.
+# Available SASL mechanisms:
+# - `PLAIN`
+# - `SCRAM-SHA-256`
+# - `SCRAM-SHA-512`
+# [wal.sasl]
+# type = "SCRAM-SHA-512"
+# username = "user_kafka"
+# password = "secret"
+
+# The Kafka TLS configuration.
+# **It's only used when the provider is `kafka`**.
+# [wal.tls]
+# server_ca_cert_path = "/path/to/server_cert"
+# client_cert_path = "/path/to/client_cert"
+# client_key_path = "/path/to/key"
+
 ## Metadata storage options.
 [metadata_store]
 ## Kv file size in bytes.
@@ -246,6 +316,7 @@ retry_delay = "500ms"
 # root = "data"
 # scope = "test"
 # credential_path = "123456"
+# credential = "base64-credential"
 # endpoint = "https://storage.googleapis.com"
 
 ## The data storage options.
@@ -261,87 +332,123 @@ data_home = "/tmp/greptimedb/"
 ## - `Oss`: the data is stored in the Aliyun OSS.
 type = "File"
 
-## Cache configuration for object storage such as 'S3' etc.
-## The local file cache directory.
-## +toml2docs:none-default
-cache_path = "/path/local_cache"
+## Read cache configuration for object storage such as 'S3' etc, it's configured by default when using object storage. It is recommended to configure it when using object storage for better performance.
+## A local file directory, defaults to `{data_home}/object_cache/read`. An empty string means disabling.
+## @toml2docs:none-default
+#+ cache_path = ""
 
-## The local file cache capacity in bytes.
-## +toml2docs:none-default
-cache_capacity = "256MB"
+## The local file cache capacity in bytes. If your disk space is sufficient, it is recommended to set it larger.
+## @toml2docs:none-default
+cache_capacity = "5GiB"
 
 ## The S3 bucket name.
 ## **It's only used when the storage type is `S3`, `Oss` and `Gcs`**.
-## +toml2docs:none-default
+## @toml2docs:none-default
 bucket = "greptimedb"
 
 ## The S3 data will be stored in the specified prefix, for example, `s3://${bucket}/${root}`.
 ## **It's only used when the storage type is `S3`, `Oss` and `Azblob`**.
-## +toml2docs:none-default
+## @toml2docs:none-default
 root = "greptimedb"
 
 ## The access key id of the aws account.
 ## It's **highly recommended** to use AWS IAM roles instead of hardcoding the access key id and secret key.
 ## **It's only used when the storage type is `S3` and `Oss`**.
-## +toml2docs:none-default
+## @toml2docs:none-default
 access_key_id = "test"
 
 ## The secret access key of the aws account.
 ## It's **highly recommended** to use AWS IAM roles instead of hardcoding the access key id and secret key.
 ## **It's only used when the storage type is `S3`**.
-## +toml2docs:none-default
+## @toml2docs:none-default
 secret_access_key = "test"
 
 ## The secret access key of the aliyun account.
 ## **It's only used when the storage type is `Oss`**.
-## +toml2docs:none-default
+## @toml2docs:none-default
 access_key_secret = "test"
 
 ## The account key of the azure account.
 ## **It's only used when the storage type is `Azblob`**.
-## +toml2docs:none-default
+## @toml2docs:none-default
 account_name = "test"
 
 ## The account key of the azure account.
 ## **It's only used when the storage type is `Azblob`**.
-## +toml2docs:none-default
+## @toml2docs:none-default
 account_key = "test"
 
 ## The scope of the google cloud storage.
 ## **It's only used when the storage type is `Gcs`**.
-## +toml2docs:none-default
+## @toml2docs:none-default
 scope = "test"
 
 ## The credential path of the google cloud storage.
 ## **It's only used when the storage type is `Gcs`**.
-## +toml2docs:none-default
+## @toml2docs:none-default
 credential_path = "test"
 
+## The credential of the google cloud storage.
+## **It's only used when the storage type is `Gcs`**.
+## @toml2docs:none-default
+credential = "base64-credential"
+
 ## The container of the azure account.
 ## **It's only used when the storage type is `Azblob`**.
-## +toml2docs:none-default
+## @toml2docs:none-default
 container = "greptimedb"
 
 ## The sas token of the azure account.
 ## **It's only used when the storage type is `Azblob`**.
-## +toml2docs:none-default
+## @toml2docs:none-default
 sas_token = ""
 
 ## The endpoint of the S3 service.
 ## **It's only used when the storage type is `S3`, `Oss`, `Gcs` and `Azblob`**.
-## +toml2docs:none-default
+## @toml2docs:none-default
 endpoint = "https://s3.amazonaws.com"
 
 ## The region of the S3 service.
 ## **It's only used when the storage type is `S3`, `Oss`, `Gcs` and `Azblob`**.
-## +toml2docs:none-default
+## @toml2docs:none-default
 region = "us-west-2"
 
+## The http client options to the storage.
+## **It's only used when the storage type is `S3`, `Oss`, `Gcs` and `Azblob`**.
+[storage.http_client]
+
+## The maximum idle connection per host allowed in the pool.
+pool_max_idle_per_host = 1024
+
+## The timeout for only the connect phase of a http client.
+connect_timeout = "30s"
+
+## The total request timeout, applied from when the request starts connecting until the response body has finished.
+## Also considered a total deadline.
+timeout = "30s"
+
+## The timeout for idle sockets being kept-alive.
+pool_idle_timeout = "90s"
+
 # Custom storage options
 # [[storage.providers]]
+# name = "S3"
 # type = "S3"
+# bucket = "greptimedb"
+# root = "data"
+# access_key_id = "test"
+# secret_access_key = "123456"
+# endpoint = "https://s3.amazonaws.com"
+# region = "us-west-2"
 # [[storage.providers]]
+# name = "Gcs"
 # type = "Gcs"
+# bucket = "greptimedb"
+# root = "data"
+# scope = "test"
+# credential_path = "123456"
+# credential = "base64-credential"
+# endpoint = "https://storage.googleapis.com"
 
 ## The region engine options. You can configure multiple region engines.
 [[region_engine]]
@@ -350,7 +457,7 @@ region = "us-west-2"
 [region_engine.mito]
 
 ## Number of region workers.
-num_workers = 8
+#+ num_workers = 8
 
 ## Request channel size of each worker.
 worker_channel_size = 128
@@ -364,57 +471,75 @@ manifest_checkpoint_distance = 10
 ## Whether to compress manifest and checkpoint file by gzip (default false).
 compress_manifest = false
 
-## Max number of running background jobs
-max_background_jobs = 4
+## Max number of running background flush jobs (default: 1/2 of cpu cores).
+## @toml2docs:none-default="Auto"
+#+ max_background_flushes = 4
+
+## Max number of running background compaction jobs (default: 1/4 of cpu cores).
+## @toml2docs:none-default="Auto"
+#+ max_background_compactions = 2
+
+## Max number of running background purge jobs (default: number of cpu cores).
+## @toml2docs:none-default="Auto"
+#+ max_background_purges = 8
 
 ## Interval to auto flush a region if it has not flushed yet.
 auto_flush_interval = "1h"
 
 ## Global write buffer size for all regions. If not set, it's default to 1/8 of OS memory with a max limitation of 1GB.
-global_write_buffer_size = "1GB"
+## @toml2docs:none-default="Auto"
+#+ global_write_buffer_size = "1GB"
 
-## Global write buffer size threshold to reject write requests. If not set, it's default to 2 times of `global_write_buffer_size`
-global_write_buffer_reject_size = "2GB"
+## Global write buffer size threshold to reject write requests. If not set, it's default to 2 times of `global_write_buffer_size`.
+## @toml2docs:none-default="Auto"
+#+ global_write_buffer_reject_size = "2GB"
 
 ## Cache size for SST metadata. Setting it to 0 to disable the cache.
 ## If not set, it's default to 1/32 of OS memory with a max limitation of 128MB.
-sst_meta_cache_size = "128MB"
+## @toml2docs:none-default="Auto"
+#+ sst_meta_cache_size = "128MB"
 
 ## Cache size for vectors and arrow arrays. Setting it to 0 to disable the cache.
 ## If not set, it's default to 1/16 of OS memory with a max limitation of 512MB.
-vector_cache_size = "512MB"
+## @toml2docs:none-default="Auto"
+#+ vector_cache_size = "512MB"
 
 ## Cache size for pages of SST row groups. Setting it to 0 to disable the cache.
-## If not set, it's default to 1/16 of OS memory with a max limitation of 512MB.
-page_cache_size = "512MB"
+## If not set, it's default to 1/8 of OS memory.
+## @toml2docs:none-default="Auto"
+#+ page_cache_size = "512MB"
 
-## Whether to enable the experimental write cache.
+## Cache size for time series selector (e.g. `last_value()`). Setting it to 0 to disable the cache.
+## If not set, it's default to 1/16 of OS memory with a max limitation of 512MB.
+## @toml2docs:none-default="Auto"
+#+ selector_result_cache_size = "512MB"
+
+## Whether to enable the experimental write cache, it's enabled by default when using object storage. It is recommended to enable it when using object storage for better performance.
 enable_experimental_write_cache = false
 
-## File system path for write cache, defaults to `{data_home}/write_cache`.
+## File system path for write cache, defaults to `{data_home}/object_cache/write`.
 experimental_write_cache_path = ""
 
-## Capacity for write cache.
-experimental_write_cache_size = "512MB"
+## Capacity for write cache. If your disk space is sufficient, it is recommended to set it larger.
+experimental_write_cache_size = "5GiB"
 
 ## TTL for write cache.
-experimental_write_cache_ttl = "1h"
+## @toml2docs:none-default
+experimental_write_cache_ttl = "8h"
 
 ## Buffer size for SST writing.
 sst_write_buffer_size = "8MB"
 
-## Parallelism to scan a region (default: 1/4 of cpu cores).
-## - `0`: using the default value (1/4 of cpu cores).
-## - `1`: scan in current thread.
-## - `n`: scan in parallelism n.
-scan_parallelism = 0
-
 ## Capacity of the channel to send data from parallel scan tasks to the main task.
 parallel_scan_channel_size = 32
 
 ## Whether to allow stale WAL entries read during replay.
 allow_stale_entries = false
 
+## Minimum time interval between two compactions.
+## To align with the old behavior, the default value is 0 (no restrictions).
+min_compaction_interval = "0m"
+
 ## The options for index in Mito engine.
 [region_engine.mito.index]
 
@@ -511,11 +636,11 @@ fork_dictionary_bytes = "1GiB"
 
 ## The logging options.
 [logging]
-## The directory to store the log files.
+## The directory to store the log files. If set to empty, logs will not be written to files.
 dir = "/tmp/greptimedb/logs"
 
 ## The log level. Can be `info`/`debug`/`warn`/`error`.
-## +toml2docs:none-default
+## @toml2docs:none-default
 level = "info"
 
 ## Enable OTLP tracing.
@@ -527,12 +652,31 @@ otlp_endpoint = "http://localhost:4317"
 ## Whether to append logs to stdout.
 append_stdout = true
 
+## The log format. Can be `text`/`json`.
+log_format = "text"
+
+## The maximum amount of log files.
+max_log_files = 720
+
 ## The percentage of tracing will be sampled and exported.
 ## Valid range `[0, 1]`, 1 means all traces are sampled, 0 means all traces are not sampled, the default value is 1.
 ## ratio > 1 are treated as 1. Fractions < 0 are treated as 0
 [logging.tracing_sample_ratio]
 default_ratio = 1.0
 
+## The slow query log options.
+[logging.slow_query]
+## Whether to enable slow query log.
+enable = false
+
+## The threshold of slow query.
+## @toml2docs:none-default
+threshold = "10s"
+
+## The sampling ratio of slow query log. The value should be in the range of (0, 1].
+## @toml2docs:none-default
+sample_ratio = 1.0
+
 ## The datanode can export its metrics and send to Prometheus compatible service (e.g. send to `greptimedb` itself) from remote-write API.
 ## This is only used for `greptimedb` to export its own metrics internally. It's different from prometheus scrape.
 [export_metrics]
@@ -543,20 +687,21 @@ enable = false
 ## The interval of export metrics.
 write_interval = "30s"
 
-## For `standalone` mode, `self_import` is recommend to collect metrics generated by itself
+## For `standalone` mode, `self_import` is recommended to collect metrics generated by itself
+## You must create the database before enabling it.
 [export_metrics.self_import]
-## +toml2docs:none-default
-db = "information_schema"
+## @toml2docs:none-default
+db = "greptime_metrics"
 
 [export_metrics.remote_write]
-## The url the metrics send to. The url example can be: `http://127.0.0.1:4000/v1/prometheus/write?db=information_schema`.
+## The url the metrics send to. The url example can be: `http://127.0.0.1:4000/v1/prometheus/write?db=greptime_metrics`.
 url = ""
 
 ## HTTP headers of Prometheus remote-write carry.
 headers = { }
 
 ## The tracing options. Only effect when compiled with `tokio-console` feature.
-[tracing]
+#+ [tracing]
 ## The tokio console address.
-## +toml2docs:none-default
-tokio_console_addr = "127.0.0.1"
+## @toml2docs:none-default
+#+ tokio_console_addr = "127.0.0.1"

docker/dev-builder/binstall/pull_binstall.sh

@@ -0,0 +1,50 @@
+#!/bin/bash
+
+set -euxo pipefail
+
+cd "$(mktemp -d)"
+# Fix version to v1.6.6, this is different than the latest version in original install script in
+# https://raw.githubusercontent.com/cargo-bins/cargo-binstall/main/install-from-binstall-release.sh
+base_url="https://github.com/cargo-bins/cargo-binstall/releases/download/v1.6.6/cargo-binstall-"
+
+os="$(uname -s)"
+if [ "$os" == "Darwin" ]; then
+    url="${base_url}universal-apple-darwin.zip"
+    curl -LO --proto '=https' --tlsv1.2 -sSf "$url"
+    unzip cargo-binstall-universal-apple-darwin.zip
+elif [ "$os" == "Linux" ]; then
+    machine="$(uname -m)"
+    if [ "$machine" == "armv7l" ]; then
+        machine="armv7"
+    fi
+    target="${machine}-unknown-linux-musl"
+    if [ "$machine" == "armv7" ]; then
+        target="${target}eabihf"
+    fi
+
+    url="${base_url}${target}.tgz"
+    curl -L --proto '=https' --tlsv1.2 -sSf "$url" | tar -xvzf -
+elif [ "${OS-}" = "Windows_NT" ]; then
+    machine="$(uname -m)"
+    target="${machine}-pc-windows-msvc"
+    url="${base_url}${target}.zip"
+    curl -LO --proto '=https' --tlsv1.2 -sSf "$url"
+    unzip "cargo-binstall-${target}.zip"
+else
+    echo "Unsupported OS ${os}"
+    exit 1
+fi
+
+./cargo-binstall -y --force cargo-binstall
+
+CARGO_HOME="${CARGO_HOME:-$HOME/.cargo}"
+
+if ! [[ ":$PATH:" == *":$CARGO_HOME/bin:"* ]]; then
+    if [ -n "${CI:-}" ] && [ -n "${GITHUB_PATH:-}" ]; then
+        echo "$CARGO_HOME/bin" >> "$GITHUB_PATH"
+    else
+        echo
+        printf "\033[0;31mYour path is missing %s, you might want to add it.\033[0m\n" "$CARGO_HOME/bin"
+        echo
+    fi
+fi

docker/dev-builder/centos/Dockerfile

@@ -32,7 +32,9 @@ RUN rustup toolchain install ${RUST_TOOLCHAIN}
 
 # Install cargo-binstall with a specific version to adapt the current rust toolchain.
 # Note: if we use the latest version, we may encounter the following `use of unstable library feature 'io_error_downcast'` error.
-RUN cargo install cargo-binstall --version 1.6.6 --locked
+# compile from source take too long, so we use the precompiled binary instead
+COPY $DOCKER_BUILD_ROOT/docker/dev-builder/binstall/pull_binstall.sh /usr/local/bin/pull_binstall.sh
+RUN chmod +x /usr/local/bin/pull_binstall.sh && /usr/local/bin/pull_binstall.sh
 
 # Install nextest.
 RUN cargo binstall cargo-nextest --no-confirm

docker/dev-builder/ubuntu/Dockerfile

@@ -24,6 +24,15 @@ RUN apt-get update && DEBIAN_FRONTEND=noninteractive apt-get install -y \
     python3.10 \
     python3.10-dev
 
+# https://github.com/GreptimeTeam/greptimedb/actions/runs/10935485852/job/30357457188#step:3:7106
+# `aws-lc-sys` require gcc >= 10.3.0 to work, hence alias to use gcc-10
+RUN apt-get remove -y gcc-9 g++-9 cpp-9 && \
+    apt-get install -y gcc-10 g++-10 cpp-10 make cmake && \
+    ln -sf /usr/bin/gcc-10 /usr/bin/gcc && ln -sf /usr/bin/g++-10 /usr/bin/g++ && \
+    ln -sf /usr/bin/gcc-10 /usr/bin/cc && \
+    ln -sf /usr/bin/g++-10 /usr/bin/cpp && ln -sf /usr/bin/g++-10 /usr/bin/c++ && \
+    cc --version && gcc --version && g++ --version && cpp --version && c++ --version
+
 # Remove Python 3.8 and install pip.
 RUN apt-get -y purge python3.8 && \
     apt-get -y autoremove && \
@@ -57,7 +66,9 @@ RUN rustup toolchain install ${RUST_TOOLCHAIN}
 
 # Install cargo-binstall with a specific version to adapt the current rust toolchain.
 # Note: if we use the latest version, we may encounter the following `use of unstable library feature 'io_error_downcast'` error.
-RUN cargo install cargo-binstall --version 1.6.6 --locked
+# compile from source take too long, so we use the precompiled binary instead
+COPY $DOCKER_BUILD_ROOT/docker/dev-builder/binstall/pull_binstall.sh /usr/local/bin/pull_binstall.sh
+RUN chmod +x /usr/local/bin/pull_binstall.sh && /usr/local/bin/pull_binstall.sh
 
 # Install nextest.
 RUN cargo binstall cargo-nextest --no-confirm

docs/benchmarks/log/README.md

@@ -0,0 +1,51 @@
+# Log benchmark configuration
+This repo holds the configuration we used to benchmark GreptimeDB, Clickhouse and Elastic Search.
+
+Here are the versions of databases we used in the benchmark
+
+| name          | version    |
+| :------------ | :--------- |
+| GreptimeDB    | v0.9.2     |
+| Clickhouse    | 24.9.1.219 |
+| Elasticsearch | 8.15.0     |
+
+## Structured model vs Unstructured model
+We divide test into two parts, using structured model and unstructured model accordingly. You can also see the difference in create table clause.
+
+__Structured model__
+
+The log data is pre-processed into columns by vector. For example an insert request looks like following
+```SQL
+INSERT INTO test_table (bytes, http_version, ip, method, path, status, user, timestamp) VALUES ()
+```
+The goal is to test string/text support for each database. In real scenarios it means the datasource(or log data producers) have separate fields defined, or have already processed the raw input.
+
+__Unstructured model__
+
+The log data is inserted as a long string, and then we build fulltext index upon these strings. For example an insert request looks like following
+```SQL
+INSERT INTO test_table (message, timestamp) VALUES ()
+```
+The goal is to test fuzzy search performance for each database. In real scenarios it means the log is produced by some kind of middleware and inserted directly into the database.
+
+## Creating tables
+See [here](./create_table.sql) for GreptimeDB and Clickhouse's create table clause.
+The mapping of Elastic search is created automatically.
+
+## Vector Configuration
+We use vector to generate random log data and send inserts to databases.
+Please refer to [structured config](./structured_vector.toml) and [unstructured config](./unstructured_vector.toml) for detailed configuration.
+
+## SQLs and payloads
+Please refer to [SQL query](./query.sql) for GreptimeDB and Clickhouse, and [query payload](./query.md) for Elastic search.
+
+## Steps to reproduce
+0. Decide whether to run structured model test or unstructured mode test.
+1. Build vector binary(see vector's config file for specific branch) and databases binaries accordingly.
+2. Create table in GreptimeDB and Clickhouse in advance.
+3. Run vector to insert data.
+4. When data insertion is finished, run queries against each database. Note: you'll need to update timerange value after data insertion.
+
+## Addition
+- You can tune GreptimeDB's configuration to get better performance.
+- You can setup GreptimeDB to use S3 as storage, see [here](https://docs.greptime.com/user-guide/deployments/configuration#storage-options).

docs/benchmarks/log/create_table.sql

@@ -0,0 +1,56 @@
+-- GreptimeDB create table clause
+-- structured test, use vector to pre-process log data into fields
+CREATE TABLE IF NOT EXISTS `test_table` (
+    `bytes` Int64 NULL,
+    `http_version` STRING NULL,
+    `ip` STRING NULL,
+    `method` STRING NULL,
+    `path` STRING NULL,
+    `status` SMALLINT UNSIGNED NULL,
+    `user` STRING NULL,
+    `timestamp` TIMESTAMP(3) NOT NULL,
+    PRIMARY KEY (`user`, `path`, `status`),
+    TIME INDEX (`timestamp`)
+)
+ENGINE=mito
+WITH(
+    append_mode = 'true'
+);
+
+-- unstructured test, build fulltext index on message column
+CREATE TABLE IF NOT EXISTS `test_table` (
+    `message` STRING NULL FULLTEXT WITH(analyzer = 'English', case_sensitive = 'false'),
+    `timestamp` TIMESTAMP(3) NOT NULL,
+    TIME INDEX (`timestamp`)
+)
+ENGINE=mito
+WITH(
+    append_mode = 'true'
+);
+
+-- Clickhouse create table clause
+-- structured test
+CREATE TABLE IF NOT EXISTS test_table
+(
+    bytes UInt64 NOT NULL,
+    http_version String NOT NULL,
+    ip String NOT NULL,
+    method String NOT NULL,
+    path String NOT NULL,
+    status UInt8 NOT NULL,
+    user String NOT NULL,
+    timestamp String NOT NULL,
+)
+ENGINE = MergeTree()
+ORDER BY (user, path, status);
+
+-- unstructured test
+SET allow_experimental_full_text_index = true;
+CREATE TABLE IF NOT EXISTS test_table
+(
+    message String,
+    timestamp String,
+    INDEX inv_idx(message) TYPE full_text(0) GRANULARITY 1
+)
+ENGINE = MergeTree()
+ORDER BY tuple();

docs/benchmarks/log/query.md

@@ -0,0 +1,199 @@
+# Query URL and payload for Elastic Search
+## Count
+URL: `http://127.0.0.1:9200/_count`
+
+## Query by timerange
+URL: `http://127.0.0.1:9200/_search`
+
+You can use the following payload to get the full timerange first.
+```JSON
+{"size":0,"aggs":{"max_timestamp":{"max":{"field":"timestamp"}},"min_timestamp":{"min":{"field":"timestamp"}}}}
+```
+
+And then use this payload to query by timerange.
+```JSON
+{
+  "from": 0,
+  "size": 1000,
+  "query": {
+    "range": {
+      "timestamp": {
+        "gte": "2024-08-16T04:30:44.000Z",
+        "lte": "2024-08-16T04:51:52.000Z"
+      }
+    }
+  }
+}
+```
+
+## Query by condition
+URL: `http://127.0.0.1:9200/_search`
+### Structured payload
+```JSON
+{
+  "from": 0,
+  "size": 10000,
+  "query": {
+    "bool": {
+      "must": [
+        {
+          "term": {
+            "user.keyword": "CrucifiX"
+          }
+        },
+        {
+          "term": {
+            "method.keyword": "OPTION"
+          }
+        },
+        {
+          "term": {
+            "path.keyword": "/user/booperbot124"
+          }
+        },
+        {
+          "term": {
+            "http_version.keyword": "HTTP/1.1"
+          }
+        },
+        {
+          "term": {
+            "status": "401"
+          }
+        }
+      ]
+    }
+  }
+}
+```
+### Unstructured payload
+```JSON
+{
+  "from": 0,
+  "size": 10000,
+  "query": {
+    "bool": {
+      "must": [
+        {
+          "match_phrase": {
+            "message": "CrucifiX"
+          }
+        },
+        {
+          "match_phrase": {
+            "message": "OPTION"
+          }
+        },
+        {
+          "match_phrase": {
+            "message": "/user/booperbot124"
+          }
+        },
+        {
+          "match_phrase": {
+            "message": "HTTP/1.1"
+          }
+        },
+        {
+          "match_phrase": {
+            "message": "401"
+          }
+        }
+      ]
+    }
+  }
+}
+```
+
+## Query by condition and timerange
+URL: `http://127.0.0.1:9200/_search`
+### Structured payload
+```JSON
+{
+  "size": 10000,
+  "query": {
+    "bool": {
+      "must": [
+        {
+          "term": {
+            "user.keyword": "CrucifiX"
+          }
+        },
+        {
+          "term": {
+            "method.keyword": "OPTION"
+          }
+        },
+        {
+          "term": {
+            "path.keyword": "/user/booperbot124"
+          }
+        },
+        {
+          "term": {
+            "http_version.keyword": "HTTP/1.1"
+          }
+        },
+        {
+          "term": {
+            "status": "401"
+          }
+        },
+        {
+          "range": {
+            "timestamp": {
+              "gte": "2024-08-19T07:03:37.383Z",
+              "lte": "2024-08-19T07:24:58.883Z"
+            }
+          }
+        }
+      ]
+    }
+  }
+}
+```
+### Unstructured payload
+```JSON
+{
+  "size": 10000,
+  "query": {
+    "bool": {
+      "must": [
+        {
+          "match_phrase": {
+            "message": "CrucifiX"
+          }
+        },
+        {
+          "match_phrase": {
+            "message": "OPTION"
+          }
+        },
+        {
+          "match_phrase": {
+            "message": "/user/booperbot124"
+          }
+        },
+        {
+          "match_phrase": {
+            "message": "HTTP/1.1"
+          }
+        },
+        {
+          "match_phrase": {
+            "message": "401"
+          }
+        },
+        {
+          "range": {
+            "timestamp": {
+              "gte": "2024-08-19T05:16:17.099Z",
+              "lte": "2024-08-19T05:46:02.722Z"
+            }
+          }
+        }
+      ]
+    }
+  }
+}
+```

docs/benchmarks/log/query.sql

@@ -0,0 +1,50 @@
+-- Structured query for GreptimeDB and Clickhouse
+
+-- query count
+select count(*) from test_table;
+
+-- query by timerange. Note: place the timestamp range in the where clause
+-- GreptimeDB
+-- you can use `select max(timestamp)::bigint from test_table;` and `select min(timestamp)::bigint from test_table;`
+-- to get the full timestamp range
+select * from test_table where timestamp between 1723710843619 and 1723711367588;
+-- Clickhouse
+-- you can use `select max(timestamp) from test_table;` and `select min(timestamp) from test_table;`
+-- to get the full timestamp range
+select * from test_table where timestamp between '2024-08-16T03:58:46Z' and '2024-08-16T04:03:50Z';
+
+-- query by condition
+SELECT * FROM test_table WHERE user = 'CrucifiX' and method = 'OPTION' and path = '/user/booperbot124' and http_version = 'HTTP/1.1' and status = 401;
+
+-- query by condition and timerange
+-- GreptimeDB
+SELECT * FROM test_table WHERE user = "CrucifiX" and method = "OPTION" and path = "/user/booperbot124" and http_version = "HTTP/1.1" and status = 401 
+and timestamp between 1723774396760 and 1723774788760;
+-- Clickhouse
+SELECT * FROM test_table WHERE user = 'CrucifiX' and method = 'OPTION' and path = '/user/booperbot124' and http_version = 'HTTP/1.1' and status = 401 
+and timestamp between '2024-08-16T03:58:46Z' and '2024-08-16T04:03:50Z';
+
+-- Unstructured query for GreptimeDB and Clickhouse
+
+
+-- query by condition
+-- GreptimeDB
+SELECT * FROM test_table WHERE MATCHES(message, "+CrucifiX +OPTION +/user/booperbot124 +HTTP/1.1 +401");
+-- Clickhouse
+SELECT * FROM test_table WHERE (message LIKE '%CrucifiX%') 
+AND (message LIKE '%OPTION%') 
+AND (message LIKE '%/user/booperbot124%') 
+AND (message LIKE '%HTTP/1.1%') 
+AND (message LIKE '%401%');
+
+-- query by condition and timerange
+-- GreptimeDB
+SELECT * FROM test_table WHERE MATCHES(message, "+CrucifiX +OPTION +/user/booperbot124 +HTTP/1.1 +401") 
+and timestamp between 1723710843619 and 1723711367588;
+-- Clickhouse
+SELECT * FROM test_table WHERE (message LIKE '%CrucifiX%') 
+AND (message LIKE '%OPTION%') 
+AND (message LIKE '%/user/booperbot124%') 
+AND (message LIKE '%HTTP/1.1%') 
+AND (message LIKE '%401%') 
+AND timestamp between '2024-08-15T10:25:26.524000000Z' AND '2024-08-15T10:31:31.746000000Z';

docs/benchmarks/log/structured_vector.toml

@@ -0,0 +1,57 @@
+# Please note we use patched branch to build vector
+# https://github.com/shuiyisong/vector/tree/chore/greptime_log_ingester_logitem
+
+[sources.demo_logs]
+type = "demo_logs"
+format = "apache_common"
+# interval value = 1 / rps
+# say you want to insert at 20k/s, that is 1 / 20000 = 0.00005
+# set to 0 to run as fast as possible
+interval = 0
+# total rows to insert
+count = 100000000
+lines = [ "line1" ]
+
+[transforms.parse_logs]
+type = "remap"
+inputs = ["demo_logs"]
+source = '''
+. = parse_regex!(.message, r'^(?P<ip>\S+) - (?P<user>\S+) \[(?P<timestamp>[^\]]+)\] "(?P<method>\S+) (?P<path>\S+) (?P<http_version>\S+)" (?P<status>\d+) (?P<bytes>\d+)$')
+
+# Convert timestamp to a standard format
+.timestamp = parse_timestamp!(.timestamp, format: "%d/%b/%Y:%H:%M:%S %z")
+
+# Convert status and bytes to integers
+.status = to_int!(.status)
+.bytes = to_int!(.bytes)
+'''
+
+[sinks.sink_greptime_logs]
+type = "greptimedb_logs"
+# The table to insert into
+table = "test_table"
+pipeline_name = "demo_pipeline"
+compression = "none"
+inputs = [ "parse_logs" ]
+endpoint = "http://127.0.0.1:4000"
+# Batch size for each insertion
+batch.max_events = 4000
+
+[sinks.clickhouse]
+type = "clickhouse"
+inputs = [ "parse_logs" ]
+database = "default"
+endpoint = "http://127.0.0.1:8123"
+format = "json_each_row"
+# The table to insert into
+table = "test_table"
+
+[sinks.sink_elasticsearch]
+type = "elasticsearch"
+inputs = [ "parse_logs" ]
+api_version = "auto"
+compression = "none"
+doc_type = "_doc"
+endpoints = [ "http://127.0.0.1:9200" ]
+id_key = "id"
+mode = "bulk"

docs/benchmarks/log/unstructured_vector.toml

@@ -0,0 +1,43 @@
+# Please note we use patched branch to build vector
+# https://github.com/shuiyisong/vector/tree/chore/greptime_log_ingester_ft
+
+[sources.demo_logs]
+type = "demo_logs"
+format = "apache_common"
+# interval value = 1 / rps
+# say you want to insert at 20k/s, that is 1 / 20000 = 0.00005
+# set to 0 to run as fast as possible
+interval = 0
+# total rows to insert
+count = 100000000
+lines = [ "line1" ]
+
+[sinks.sink_greptime_logs]
+type = "greptimedb_logs"
+# The table to insert into
+table = "test_table"
+pipeline_name = "demo_pipeline"
+compression = "none"
+inputs = [ "demo_logs" ]
+endpoint = "http://127.0.0.1:4000"
+# Batch size for each insertion
+batch.max_events = 500
+
+[sinks.clickhouse]
+type = "clickhouse"
+inputs = [ "demo_logs" ]
+database = "default"
+endpoint = "http://127.0.0.1:8123"
+format = "json_each_row"
+# The table to insert into
+table = "test_table"
+
+[sinks.sink_elasticsearch]
+type = "elasticsearch"
+inputs = [ "demo_logs" ]
+api_version = "auto"
+compression = "none"
+doc_type = "_doc"
+endpoints = [ "http://127.0.0.1:9200" ]
+id_key = "id"
+mode = "bulk"

docs/how-to/how-to-change-log-level-on-the-fly.md

@@ -0,0 +1,16 @@
+# Change Log Level on the Fly
+
+## HTTP API
+
+example:
+```bash
+curl --data "trace,flow=debug" 127.0.0.1:4000/debug/log_level
+```
+And database will reply with something like:
+```bash
+Log Level changed from Some("info") to "trace,flow=debug"%
+```
+
+The data is a string in the format of `global_level,module1=level1,module2=level2,...` that follow the same rule of `RUST_LOG`. 
+
+The module is the module name of the log, and the level is the log level. The log level can be one of the following: `trace`, `debug`, `info`, `warn`, `error`, `off`(case insensitive).

docs/how-to/how-to-profile-cpu.md

@@ -1,15 +1,9 @@
 # Profiling CPU
 
-## Build GreptimeDB with `pprof` feature
-
-```bash
-cargo build --features=pprof
-```
-
 ## HTTP API
 Sample at 99 Hertz, for 5 seconds, output report in [protobuf format](https://github.com/google/pprof/blob/master/proto/profile.proto).
 ```bash
-curl -s '0:4000/v1/prof/cpu' > /tmp/pprof.out
+curl -X POST -s '0:4000/debug/prof/cpu' > /tmp/pprof.out
 ```
 
 Then you can use `pprof` command with the protobuf file.
@@ -19,10 +13,10 @@ go tool pprof -top /tmp/pprof.out
 
 Sample at 99 Hertz, for 60 seconds, output report in flamegraph format.
 ```bash
-curl -s '0:4000/v1/prof/cpu?seconds=60&output=flamegraph' > /tmp/pprof.svg
+curl -X POST -s '0:4000/debug/prof/cpu?seconds=60&output=flamegraph' > /tmp/pprof.svg
 ```
 
 Sample at 49 Hertz, for 10 seconds, output report in text format.
 ```bash
-curl -s '0:4000/v1/prof/cpu?seconds=10&frequency=49&output=text' > /tmp/pprof.txt
+curl -X POST -s '0:4000/debug/prof/cpu?seconds=10&frequency=49&output=text' > /tmp/pprof.txt
 ```

docs/how-to/how-to-profile-memory.md

@@ -12,16 +12,10 @@ brew install jemalloc
 sudo apt install libjemalloc-dev
 ```
 
-### [flamegraph](https://github.com/brendangregg/FlameGraph) 
+### [flamegraph](https://github.com/brendangregg/FlameGraph)
 
 ```bash
-curl https://raw.githubusercontent.com/brendangregg/FlameGraph/master/flamegraph.pl > ./flamegraph.pl 
-```
-
-### Build GreptimeDB with `mem-prof` feature.
-
-```bash
-cargo build --features=mem-prof
+curl https://raw.githubusercontent.com/brendangregg/FlameGraph/master/flamegraph.pl > ./flamegraph.pl
 ```
 
 ## Profiling
@@ -29,13 +23,13 @@ cargo build --features=mem-prof
 Start GreptimeDB instance with environment variables:
 
 ```bash
-MALLOC_CONF=prof:true,lg_prof_interval:28 ./target/debug/greptime standalone start
+MALLOC_CONF=prof:true ./target/debug/greptime standalone start
 ```
 
 Dump memory profiling data through HTTP API:
 
 ```bash
-curl localhost:4000/v1/prof/mem > greptime.hprof
+curl -X POST localhost:4000/debug/prof/mem > greptime.hprof
 ```
 
 You can periodically dump profiling data and compare them to find the delta memory usage.
@@ -45,6 +39,9 @@ You can periodically dump profiling data and compare them to find the delta memo
 To create flamegraph according to dumped profiling data:
 
 ```bash
-jeprof --svg <path_to_greptimedb_binary> --base=<baseline_prof> <profile_data> > output.svg
-```
+sudo apt install -y libjemalloc-dev
 
+jeprof <path_to_greptime_binary> <profile_data> --collapse | ./flamegraph.pl > mem-prof.svg
+
+jeprof <path_to_greptime_binary> --base <baseline_prof> <profile_data> --collapse | ./flamegraph.pl > output.svg
+```

docs/rfcs/2024-08-06-json-datatype.md

@@ -0,0 +1,197 @@
+---
+Feature Name: Json Datatype
+Tracking Issue: https://github.com/GreptimeTeam/greptimedb/issues/4230
+Date: 2024-8-6
+Author: "Yuhan Wang <profsyb@gmail.com>"
+---
+
+# Summary
+This RFC proposes a method for storing and querying JSON data in the database.
+
+# Motivation
+JSON is widely used across various scenarios. Direct support for writing and querying JSON can significantly enhance the database's flexibility.
+
+# Details
+
+## Storage and Query
+
+GreptimeDB's type system is built on Arrow/DataFusion, where each data type in GreptimeDB corresponds to a data type in Arrow/DataFusion. The proposed JSON type will be implemented on top of the existing `Binary` type, leveraging the current `datatype::value::Value` and `datatype::vectors::BinaryVector` implementations, utilizing the JSONB format as the encoding of JSON data. JSON data is stored and processed similarly to binary data within the storage layer and query engine.
+
+This approach brings problems when dealing with insertions and queries of JSON columns.
+
+## Insertion
+
+Users commonly write JSON data as strings. Thus we need to make conversions between string and JSONB. There are 2 ways to do this:
+
+1. MySQL and PostgreSQL servers provide auto-conversions between strings and JSONB. When a string is inserted into a JSON column, the server will try to parse the string as JSON and convert it to JSONB. The non-JSON strings will be rejected.
+
+2. A function `parse_json` is provided to convert string to JSONB. If the string is not a valid JSON string, the function will return an error.
+
+For example, in MySQL client:
+```SQL
+CREATE TABLE IF NOT EXISTS test (
+    ts TIMESTAMP TIME INDEX,
+    a INT,
+    b JSON
+);
+
+INSERT INTO test VALUES(
+    0,
+    0,
+    '{
+        "name": "jHl2oDDnPc1i2OzlP5Y",
+        "timestamp": "2024-07-25T04:33:11.369386Z",
+        "attributes": { "event_attributes": 48.28667 }
+    }'
+);
+
+INSERT INTO test VALUES(
+    0,
+    0,
+    parse_json('{
+        "name": "jHl2oDDnPc1i2OzlP5Y",
+        "timestamp": "2024-07-25T04:33:11.369386Z",
+        "attributes": { "event_attributes": 48.28667 }
+    }')
+);
+```
+Are both valid.
+
+The dataflow of the insertion process is as follows:
+```
+Insert JSON strings directly through client:
+                                   Parse                       Insert
+        String(Serialized JSON)┌──────────┐Arrow Binary(JSONB)┌──────┐Arrow Binary(JSONB)
+ Client ---------------------->│  Server  │------------------>│ Mito │------------------> Storage
+                               └──────────┘                   └──────┘
+        (Server identifies JSON type and performs auto-conversion)
+
+Insert JSON strings through parse_json function:
+                                                                   Parse                     Insert
+        String(Serialized JSON)┌──────────┐String(Serialized JSON)┌─────┐Arrow Binary(JSONB)┌──────┐Arrow Binary(JSONB)
+ Client ---------------------->│  Server  │---------------------->│ UDF │------------------>│ Mito │------------------> Storage
+                               └──────────┘                       └─────┘                   └──────┘
+                                            (Conversion is performed by UDF inside Query Engine)
+```
+
+Servers identify JSON column through column schema and perform auto-conversions. But when using prepared statements and binding parameters, the corresponding cached plans in datafusion generated by prepared statements cannot identify JSON columns. Under this circumstance, the servers identify JSON columns through the given parameters and perform auto-conversions.
+
+The following is an example of inserting JSON data through prepared statements:
+```Rust
+sqlx::query(
+    "create table test(ts timestamp time index, j json)",
+)
+.execute(&pool)
+.await
+.unwrap();
+
+let json = serde_json::json!({
+    "code": 200,
+    "success": true,
+    "payload": {
+        "features": [
+            "serde",
+            "json"
+        ],
+        "homepage": null
+    }
+});
+
+// Valid, can identify serde_json::Value as JSON type
+sqlx::query("insert into test values($1, $2)")
+    .bind(i)
+    .bind(json)
+    .execute(&pool)
+    .await
+    .unwrap();
+
+// Invalid, cannot identify String as JSON type
+sqlx::query("insert into test values($1, $2)")
+    .bind(i)
+    .bind(json.to_string())
+    .execute(&pool)
+    .await
+    .unwrap();
+```
+
+## Query
+
+Correspondingly, users prefer to display JSON data as strings. Thus we need to make conversions between JSON data and strings before presenting JSON data. There are also 2 ways to do this: auto-conversions on MySQL and PostgreSQL servers, and function `json_to_string`.
+
+For example, in MySQL client:
+```SQL
+SELECT b FROM test;
+
+SELECT json_to_string(b) FROM test;
+```
+Will both return the JSON as human-readable strings.
+
+Specifically, to perform auto-conversions, we attach a message to JSON data in the `metadata` of `Field` in Arrow/Datafusion schema when scanning a JSON column. Frontend servers could identify JSON data and convert it to strings.
+
+The dataflow of the query process is as follows:
+```
+Query directly through client:
+                                  Decode                            Scan
+        String(Serialized JSON)┌──────────┐Arrow Binary(JSONB)┌──────────────┐Arrow Binary(JSONB)
+ Client <----------------------│  Server  │<------------------│ Query Engine │<----------------- Storage
+                               └──────────┘                   └──────────────┘
+(Server identifies JSON type and performs auto-conversion based on column metadata)
+
+Query through json_to_string function:
+                                                                   Scan & Decode
+        String(Serialized JSON)┌──────────┐String(Serialized JSON)┌──────────────┐Arrow Binary(JSONB)
+ Client <----------------------│  Server  │<----------------------│ Query Engine │<----------------- Storage
+                               └──────────┘                       └──────────────┘
+                                                 (Conversion is performed by UDF inside Query Engine)
+
+```
+
+However, if a function uses JSON type as its return type, the metadata method mentioned above is not applicable. Thus the functions of JSON type should specify the return type explicitly instead of returning a JSON type, such as `json_get_int` and `json_get_float` which return corresponding data of `INT` and `FLOAT` type respectively.
+
+## Functions
+Similar to the common JSON type, JSON data can be queried with functions.
+
+For example:
+```SQL
+CREATE TABLE IF NOT EXISTS test (
+    ts TIMESTAMP TIME INDEX,
+    a INT,
+    b JSON
+);
+
+INSERT INTO test VALUES(
+    0,
+    0,
+    '{
+        "name": "jHl2oDDnPc1i2OzlP5Y",
+        "timestamp": "2024-07-25T04:33:11.369386Z",
+        "attributes": { "event_attributes": 48.28667 }
+    }'
+);
+
+SELECT json_get_string(b, 'name') FROM test;
++---------------------+
+| b.name              |
++---------------------+
+| jHl2oDDnPc1i2OzlP5Y |
++---------------------+
+
+SELECT json_get_float(b, 'attributes.event_attributes') FROM test;
++--------------------------------+
+| b.attributes.event_attributes  |
++--------------------------------+
+| 48.28667                       |
++--------------------------------+
+
+```
+And more functions can be added in the future.
+
+# Drawbacks
+
+As a general purpose JSON data type, JSONB may not be as efficient as specialized data types for specific scenarios.
+
+The auto-conversion mechanism is not supported in all scenarios. We need to find workarounds for these scenarios.
+
+# Alternatives
+
+Extract and flatten JSON schema to store in a structured format through pipeline. For nested data, we can provide nested types like `STRUCT` or `ARRAY`.

grafana/README.md

@@ -5,6 +5,13 @@ GreptimeDB's official Grafana dashboard.
 
 Status notify: we are still working on this config. It's expected to change frequently in the recent days. Please feel free to submit your feedback and/or contribution to this dashboard 🤗
 
+If you use Helm [chart](https://github.com/GreptimeTeam/helm-charts) to deploy GreptimeDB cluster, you can enable self-monitoring by setting the following values in your Helm chart:
+
+- `monitoring.enabled=true`: Deploys a standalone GreptimeDB instance dedicated to monitoring the cluster;
+- `grafana.enabled=true`: Deploys Grafana and automatically imports the monitoring dashboard;
+
+The standalone GreptimeDB instance will collect metrics from your cluster and the dashboard will be available in the Grafana UI. For detailed deployment instructions, please refer to our [Kubernetes deployment guide](https://docs.greptime.com/nightly/user-guide/deployments/deploy-on-kubernetes/getting-started).
+
 # How to use
 
 ## `greptimedb.json`
@@ -25,7 +32,7 @@ Please ensure the following configuration before importing the dashboard into Gr
 
 __1. Prometheus scrape config__
 
-Assign `greptime_pod` label to each host target. We use this label to identify each node instance.
+Configure Prometheus to scrape the cluster.
 
 ```yml
 # example config
@@ -34,27 +41,15 @@ Assign `greptime_pod` label to each host target. We use this label to identify e
 scrape_configs:
   - job_name: metasrv
     static_configs:
-    - targets: ['<ip>:<port>']
-      labels:
-        greptime_pod: metasrv
+    - targets: ['<metasrv-ip>:<port>']
 
   - job_name: datanode
     static_configs:
-    - targets: ['<ip>:<port>']
-      labels:
-        greptime_pod: datanode1
-    - targets: ['<ip>:<port>']
-      labels:
-        greptime_pod: datanode2
-    - targets: ['<ip>:<port>']
-      labels:
-        greptime_pod: datanode3
+    - targets: ['<datanode0-ip>:<port>', '<datanode1-ip>:<port>', '<datanode2-ip>:<port>']
 
   - job_name: frontend
     static_configs:
-    - targets: ['<ip>:<port>']
-      labels:
-        greptime_pod: frontend
+    - targets: ['<frontend-ip>:<port>']
 ```
 
 __2. Grafana config__
@@ -63,4 +58,4 @@ Create a Prometheus data source in Grafana before using this dashboard. We use `
 
 ### Usage
 
-Use `datasource` or `greptime_pod` on the upper-left corner to filter data from certain node.
+Use `datasource` or `instance` on the upper-left corner to filter data from certain node.

rust-toolchain.toml

@@ -1,2 +1,2 @@
 [toolchain]
-channel = "nightly-2024-04-20"
+channel = "nightly-2024-10-19"

scripts/check-builder-rust-version.sh

@@ -0,0 +1,42 @@
+#!/usr/bin/env bash
+
+set -e
+
+RUST_TOOLCHAIN_VERSION_FILE="rust-toolchain.toml"
+DEV_BUILDER_UBUNTU_REGISTRY="docker.io"
+DEV_BUILDER_UBUNTU_NAMESPACE="greptime"
+DEV_BUILDER_UBUNTU_NAME="dev-builder-ubuntu"
+
+function check_rust_toolchain_version() {
+  DEV_BUILDER_IMAGE_TAG=$(grep "DEV_BUILDER_IMAGE_TAG ?= " Makefile | cut -d= -f2 | sed 's/^[ \t]*//')
+  if [ -z "$DEV_BUILDER_IMAGE_TAG" ]; then
+    echo "Error: No DEV_BUILDER_IMAGE_TAG found in Makefile"
+    exit 1
+  fi
+
+  DEV_BUILDER_UBUNTU_IMAGE="$DEV_BUILDER_UBUNTU_REGISTRY/$DEV_BUILDER_UBUNTU_NAMESPACE/$DEV_BUILDER_UBUNTU_NAME:$DEV_BUILDER_IMAGE_TAG"
+
+  CURRENT_VERSION=$(grep -Eo '[0-9]{4}-[0-9]{2}-[0-9]{2}' "$RUST_TOOLCHAIN_VERSION_FILE")
+  if [ -z "$CURRENT_VERSION" ]; then
+    echo "Error: No rust toolchain version found in $RUST_TOOLCHAIN_VERSION_FILE"
+    exit 1
+  fi
+
+  RUST_TOOLCHAIN_VERSION_IN_BUILDER=$(docker run "$DEV_BUILDER_UBUNTU_IMAGE" rustc --version | grep  -Eo '[0-9]{4}-[0-9]{2}-[0-9]{2}')
+  if [ -z "$RUST_TOOLCHAIN_VERSION_IN_BUILDER" ]; then
+    echo "Error: No rustc version found in $DEV_BUILDER_UBUNTU_IMAGE"
+    exit 1
+  fi
+
+  # Compare the version and the difference should be less than 1 day.
+  current_rust_toolchain_seconds=$(date -d "$CURRENT_VERSION" +%s)
+  rust_toolchain_in_dev_builder_ubuntu_seconds=$(date -d "$RUST_TOOLCHAIN_VERSION_IN_BUILDER" +%s)
+  date_diff=$(( (current_rust_toolchain_seconds - rust_toolchain_in_dev_builder_ubuntu_seconds) / 86400 ))
+
+  if [ $date_diff -gt 1 ]; then
+    echo "Error: The rust toolchain '$RUST_TOOLCHAIN_VERSION_IN_BUILDER' in builder '$DEV_BUILDER_UBUNTU_IMAGE' maybe outdated, please update it to '$CURRENT_VERSION'"
+    exit 1
+  fi
+}
+
+check_rust_toolchain_version

scripts/check-snafu.py

@@ -0,0 +1,69 @@
+# Copyright 2023 Greptime Team
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+import os
+import re
+
+
+def find_rust_files(directory):
+    error_files = []
+    other_rust_files = []
+    for root, _, files in os.walk(directory):
+        for file in files:
+            if file == "error.rs":
+                error_files.append(os.path.join(root, file))
+            elif file.endswith(".rs"):
+                other_rust_files.append(os.path.join(root, file))
+    return error_files, other_rust_files
+
+
+def extract_branch_names(file_content):
+    pattern = re.compile(r"#\[snafu\(display\([^\)]*\)\)\]\s*(\w+)\s*\{")
+    return pattern.findall(file_content)
+
+
+def check_snafu_in_files(branch_name, rust_files):
+    branch_name_snafu = f"{branch_name}Snafu"
+    for rust_file in rust_files:
+        with open(rust_file, "r") as file:
+            content = file.read()
+            if branch_name_snafu in content:
+                return True
+    return False
+
+
+def main():
+    error_files, other_rust_files = find_rust_files(".")
+    branch_names = []
+
+    for error_file in error_files:
+        with open(error_file, "r") as file:
+            content = file.read()
+            branch_names.extend(extract_branch_names(content))
+
+    unused_snafu = [
+        branch_name
+        for branch_name in branch_names
+        if not check_snafu_in_files(branch_name, other_rust_files)
+    ]
+
+    for name in unused_snafu:
+        print(name)
+
+    if unused_snafu:
+        raise SystemExit(1)
+
+
+if __name__ == "__main__":
+    main()

scripts/install.sh

@@ -1,4 +1,4 @@
-#!/usr/bin/env bash
+#!/bin/sh
 
 set -ue
 
@@ -15,7 +15,7 @@ GITHUB_ORG=GreptimeTeam
 GITHUB_REPO=greptimedb
 BIN=greptime
 
-function get_os_type() {
+get_os_type() {
   os_type="$(uname -s)"
 
   case "$os_type" in
@@ -31,7 +31,7 @@ function get_os_type() {
   esac
 }
 
-function get_arch_type() {
+get_arch_type() {
   arch_type="$(uname -m)"
 
   case "$arch_type" in
@@ -53,7 +53,7 @@ function get_arch_type() {
   esac
 }
 
-function download_artifact() {
+download_artifact() {
   if [ -n "${OS_TYPE}" ] && [ -n "${ARCH_TYPE}" ]; then
     # Use the latest stable released version.
     # GitHub API reference: https://docs.github.com/en/rest/releases/releases?apiVersion=2022-11-28#get-the-latest-release.

src/api/src/helper.rs

@@ -17,10 +17,11 @@ use std::sync::Arc;
 use common_base::BitVec;
 use common_decimal::decimal128::{DECIMAL128_DEFAULT_SCALE, DECIMAL128_MAX_PRECISION};
 use common_decimal::Decimal128;
-use common_time::interval::IntervalUnit;
 use common_time::time::Time;
 use common_time::timestamp::TimeUnit;
-use common_time::{Date, DateTime, Interval, Timestamp};
+use common_time::{
+    Date, DateTime, IntervalDayTime, IntervalMonthDayNano, IntervalYearMonth, Timestamp,
+};
 use datatypes::prelude::{ConcreteDataType, ValueRef};
 use datatypes::scalars::ScalarVector;
 use datatypes::types::{
@@ -35,14 +36,14 @@ use datatypes::vectors::{
     TimestampMillisecondVector, TimestampNanosecondVector, TimestampSecondVector, UInt32Vector,
     UInt64Vector, VectorRef,
 };
-use greptime_proto::v1;
 use greptime_proto::v1::column_data_type_extension::TypeExt;
 use greptime_proto::v1::ddl_request::Expr;
 use greptime_proto::v1::greptime_request::Request;
 use greptime_proto::v1::query_request::Query;
 use greptime_proto::v1::value::ValueData;
 use greptime_proto::v1::{
-    ColumnDataTypeExtension, DdlRequest, DecimalTypeExtension, QueryRequest, Row, SemanticType,
+    self, ColumnDataTypeExtension, DdlRequest, DecimalTypeExtension, JsonTypeExtension,
+    QueryRequest, Row, SemanticType, VectorTypeExtension,
 };
 use paste::paste;
 use snafu::prelude::*;
@@ -103,7 +104,18 @@ impl From<ColumnDataTypeWrapper> for ConcreteDataType {
             ColumnDataType::Uint64 => ConcreteDataType::uint64_datatype(),
             ColumnDataType::Float32 => ConcreteDataType::float32_datatype(),
             ColumnDataType::Float64 => ConcreteDataType::float64_datatype(),
-            ColumnDataType::Binary => ConcreteDataType::binary_datatype(),
+            ColumnDataType::Binary => {
+                if let Some(TypeExt::JsonType(_)) = datatype_wrapper
+                    .datatype_ext
+                    .as_ref()
+                    .and_then(|datatype_ext| datatype_ext.type_ext.as_ref())
+                {
+                    ConcreteDataType::json_datatype()
+                } else {
+                    ConcreteDataType::binary_datatype()
+                }
+            }
+            ColumnDataType::Json => ConcreteDataType::json_datatype(),
             ColumnDataType::String => ConcreteDataType::string_datatype(),
             ColumnDataType::Date => ConcreteDataType::date_datatype(),
             ColumnDataType::Datetime => ConcreteDataType::datetime_datatype(),
@@ -137,6 +149,17 @@ impl From<ColumnDataTypeWrapper> for ConcreteDataType {
                     ConcreteDataType::decimal128_default_datatype()
                 }
             }
+            ColumnDataType::Vector => {
+                if let Some(TypeExt::VectorType(d)) = datatype_wrapper
+                    .datatype_ext
+                    .as_ref()
+                    .and_then(|datatype_ext| datatype_ext.type_ext.as_ref())
+                {
+                    ConcreteDataType::vector_datatype(d.dim)
+                } else {
+                    ConcreteDataType::vector_default_datatype()
+                }
+            }
         }
     }
 }
@@ -218,6 +241,15 @@ impl ColumnDataTypeWrapper {
             }),
         }
     }
+
+    pub fn vector_datatype(dim: u32) -> Self {
+        ColumnDataTypeWrapper {
+            datatype: ColumnDataType::Vector,
+            datatype_ext: Some(ColumnDataTypeExtension {
+                type_ext: Some(TypeExt::VectorType(VectorTypeExtension { dim })),
+            }),
+        }
+    }
 }
 
 impl TryFrom<ConcreteDataType> for ColumnDataTypeWrapper {
@@ -258,6 +290,8 @@ impl TryFrom<ConcreteDataType> for ColumnDataTypeWrapper {
                 IntervalType::MonthDayNano(_) => ColumnDataType::IntervalMonthDayNano,
             },
             ConcreteDataType::Decimal128(_) => ColumnDataType::Decimal128,
+            ConcreteDataType::Json(_) => ColumnDataType::Json,
+            ConcreteDataType::Vector(_) => ColumnDataType::Vector,
             ConcreteDataType::Null(_)
             | ConcreteDataType::List(_)
             | ConcreteDataType::Dictionary(_)
@@ -276,6 +310,18 @@ impl TryFrom<ConcreteDataType> for ColumnDataTypeWrapper {
                         })),
                     })
             }
+            ColumnDataType::Json => datatype.as_json().map(|_| ColumnDataTypeExtension {
+                type_ext: Some(TypeExt::JsonType(JsonTypeExtension::JsonBinary.into())),
+            }),
+            ColumnDataType::Vector => {
+                datatype
+                    .as_vector()
+                    .map(|vector_type| ColumnDataTypeExtension {
+                        type_ext: Some(TypeExt::VectorType(VectorTypeExtension {
+                            dim: vector_type.dim as _,
+                        })),
+                    })
+            }
             _ => None,
         };
         Ok(Self {
@@ -395,6 +441,14 @@ pub fn values_with_capacity(datatype: ColumnDataType, capacity: usize) -> Values
             decimal128_values: Vec::with_capacity(capacity),
             ..Default::default()
         },
+        ColumnDataType::Json => Values {
+            string_values: Vec::with_capacity(capacity),
+            ..Default::default()
+        },
+        ColumnDataType::Vector => Values {
+            binary_values: Vec::with_capacity(capacity),
+            ..Default::default()
+        },
     }
 }
 
@@ -435,13 +489,11 @@ pub fn push_vals(column: &mut Column, origin_count: usize, vector: VectorRef) {
             TimeUnit::Microsecond => values.time_microsecond_values.push(val.value()),
             TimeUnit::Nanosecond => values.time_nanosecond_values.push(val.value()),
         },
-        Value::Interval(val) => match val.unit() {
-            IntervalUnit::YearMonth => values.interval_year_month_values.push(val.to_i32()),
-            IntervalUnit::DayTime => values.interval_day_time_values.push(val.to_i64()),
-            IntervalUnit::MonthDayNano => values
-                .interval_month_day_nano_values
-                .push(convert_i128_to_interval(val.to_i128())),
-        },
+        Value::IntervalYearMonth(val) => values.interval_year_month_values.push(val.to_i32()),
+        Value::IntervalDayTime(val) => values.interval_day_time_values.push(val.to_i64()),
+        Value::IntervalMonthDayNano(val) => values
+            .interval_month_day_nano_values
+            .push(convert_month_day_nano_to_pb(val)),
         Value::Decimal128(val) => values.decimal128_values.push(convert_to_pb_decimal128(val)),
         Value::List(_) | Value::Duration(_) => unreachable!(),
     });
@@ -475,25 +527,24 @@ fn ddl_request_type(request: &DdlRequest) -> &'static str {
     match request.expr {
         Some(Expr::CreateDatabase(_)) => "ddl.create_database",
         Some(Expr::CreateTable(_)) => "ddl.create_table",
-        Some(Expr::Alter(_)) => "ddl.alter",
+        Some(Expr::AlterTable(_)) => "ddl.alter_table",
         Some(Expr::DropTable(_)) => "ddl.drop_table",
         Some(Expr::TruncateTable(_)) => "ddl.truncate_table",
         Some(Expr::CreateFlow(_)) => "ddl.create_flow",
         Some(Expr::DropFlow(_)) => "ddl.drop_flow",
         Some(Expr::CreateView(_)) => "ddl.create_view",
         Some(Expr::DropView(_)) => "ddl.drop_view",
+        Some(Expr::AlterDatabase(_)) => "ddl.alter_database",
         None => "ddl.empty",
     }
 }
 
-/// Converts an i128 value to google protobuf type [IntervalMonthDayNano].
-pub fn convert_i128_to_interval(v: i128) -> v1::IntervalMonthDayNano {
-    let interval = Interval::from_i128(v);
-    let (months, days, nanoseconds) = interval.to_month_day_nano();
+/// Converts an interval to google protobuf type [IntervalMonthDayNano].
+pub fn convert_month_day_nano_to_pb(v: IntervalMonthDayNano) -> v1::IntervalMonthDayNano {
     v1::IntervalMonthDayNano {
-        months,
-        days,
-        nanoseconds,
+        months: v.months,
+        days: v.days,
+        nanoseconds: v.nanoseconds,
     }
 }
 
@@ -541,11 +592,15 @@ pub fn pb_value_to_value_ref<'a>(
         ValueData::TimeMillisecondValue(t) => ValueRef::Time(Time::new_millisecond(*t)),
         ValueData::TimeMicrosecondValue(t) => ValueRef::Time(Time::new_microsecond(*t)),
         ValueData::TimeNanosecondValue(t) => ValueRef::Time(Time::new_nanosecond(*t)),
-        ValueData::IntervalYearMonthValue(v) => ValueRef::Interval(Interval::from_i32(*v)),
-        ValueData::IntervalDayTimeValue(v) => ValueRef::Interval(Interval::from_i64(*v)),
+        ValueData::IntervalYearMonthValue(v) => {
+            ValueRef::IntervalYearMonth(IntervalYearMonth::from_i32(*v))
+        }
+        ValueData::IntervalDayTimeValue(v) => {
+            ValueRef::IntervalDayTime(IntervalDayTime::from_i64(*v))
+        }
         ValueData::IntervalMonthDayNanoValue(v) => {
-            let interval = Interval::from_month_day_nano(v.months, v.days, v.nanoseconds);
-            ValueRef::Interval(interval)
+            let interval = IntervalMonthDayNano::new(v.months, v.days, v.nanoseconds);
+            ValueRef::IntervalMonthDayNano(interval)
         }
         ValueData::Decimal128Value(v) => {
             // get precision and scale from datatype_extension
@@ -636,7 +691,7 @@ pub fn pb_values_to_vector_ref(data_type: &ConcreteDataType, values: Values) ->
             IntervalType::MonthDayNano(_) => {
                 Arc::new(IntervalMonthDayNanoVector::from_iter_values(
                     values.interval_month_day_nano_values.iter().map(|x| {
-                        Interval::from_month_day_nano(x.months, x.days, x.nanoseconds).to_i128()
+                        IntervalMonthDayNano::new(x.months, x.days, x.nanoseconds).to_i128()
                     }),
                 ))
             }
@@ -646,10 +701,12 @@ pub fn pb_values_to_vector_ref(data_type: &ConcreteDataType, values: Values) ->
                 Decimal128::from_value_precision_scale(x.hi, x.lo, d.precision(), d.scale()).into()
             }),
         )),
+        ConcreteDataType::Vector(_) => Arc::new(BinaryVector::from_vec(values.binary_values)),
         ConcreteDataType::Null(_)
         | ConcreteDataType::List(_)
         | ConcreteDataType::Dictionary(_)
-        | ConcreteDataType::Duration(_) => {
+        | ConcreteDataType::Duration(_)
+        | ConcreteDataType::Json(_) => {
             unreachable!()
         }
     }
@@ -780,18 +837,18 @@ pub fn pb_values_to_values(data_type: &ConcreteDataType, values: Values) -> Vec<
         ConcreteDataType::Interval(IntervalType::YearMonth(_)) => values
             .interval_year_month_values
             .into_iter()
-            .map(|v| Value::Interval(Interval::from_i32(v)))
+            .map(|v| Value::IntervalYearMonth(IntervalYearMonth::from_i32(v)))
             .collect(),
         ConcreteDataType::Interval(IntervalType::DayTime(_)) => values
             .interval_day_time_values
             .into_iter()
-            .map(|v| Value::Interval(Interval::from_i64(v)))
+            .map(|v| Value::IntervalDayTime(IntervalDayTime::from_i64(v)))
             .collect(),
         ConcreteDataType::Interval(IntervalType::MonthDayNano(_)) => values
             .interval_month_day_nano_values
             .into_iter()
             .map(|v| {
-                Value::Interval(Interval::from_month_day_nano(
+                Value::IntervalMonthDayNano(IntervalMonthDayNano::new(
                     v.months,
                     v.days,
                     v.nanoseconds,
@@ -810,10 +867,12 @@ pub fn pb_values_to_values(data_type: &ConcreteDataType, values: Values) -> Vec<
                 ))
             })
             .collect(),
+        ConcreteDataType::Vector(_) => values.binary_values.into_iter().map(|v| v.into()).collect(),
         ConcreteDataType::Null(_)
         | ConcreteDataType::List(_)
         | ConcreteDataType::Dictionary(_)
-        | ConcreteDataType::Duration(_) => {
+        | ConcreteDataType::Duration(_)
+        | ConcreteDataType::Json(_) => {
             unreachable!()
         }
     }
@@ -831,7 +890,10 @@ pub fn is_column_type_value_eq(
     expect_type: &ConcreteDataType,
 ) -> bool {
     ColumnDataTypeWrapper::try_new(type_value, type_extension)
-        .map(|wrapper| ConcreteDataType::from(wrapper) == *expect_type)
+        .map(|wrapper| {
+            let datatype = ConcreteDataType::from(wrapper);
+            expect_type == &datatype
+        })
         .unwrap_or(false)
 }
 
@@ -912,18 +974,16 @@ pub fn to_proto_value(value: Value) -> Option<v1::Value> {
                 value_data: Some(ValueData::TimeNanosecondValue(v.value())),
             },
         },
-        Value::Interval(v) => match v.unit() {
-            IntervalUnit::YearMonth => v1::Value {
-                value_data: Some(ValueData::IntervalYearMonthValue(v.to_i32())),
-            },
-            IntervalUnit::DayTime => v1::Value {
-                value_data: Some(ValueData::IntervalDayTimeValue(v.to_i64())),
-            },
-            IntervalUnit::MonthDayNano => v1::Value {
-                value_data: Some(ValueData::IntervalMonthDayNanoValue(
-                    convert_i128_to_interval(v.to_i128()),
-                )),
-            },
+        Value::IntervalYearMonth(v) => v1::Value {
+            value_data: Some(ValueData::IntervalYearMonthValue(v.to_i32())),
+        },
+        Value::IntervalDayTime(v) => v1::Value {
+            value_data: Some(ValueData::IntervalDayTimeValue(v.to_i64())),
+        },
+        Value::IntervalMonthDayNano(v) => v1::Value {
+            value_data: Some(ValueData::IntervalMonthDayNanoValue(
+                convert_month_day_nano_to_pb(v),
+            )),
         },
         Value::Decimal128(v) => v1::Value {
             value_data: Some(ValueData::Decimal128Value(convert_to_pb_decimal128(v))),
@@ -1015,13 +1075,11 @@ pub fn value_to_grpc_value(value: Value) -> GrpcValue {
                 TimeUnit::Microsecond => ValueData::TimeMicrosecondValue(v.value()),
                 TimeUnit::Nanosecond => ValueData::TimeNanosecondValue(v.value()),
             }),
-            Value::Interval(v) => Some(match v.unit() {
-                IntervalUnit::YearMonth => ValueData::IntervalYearMonthValue(v.to_i32()),
-                IntervalUnit::DayTime => ValueData::IntervalDayTimeValue(v.to_i64()),
-                IntervalUnit::MonthDayNano => {
-                    ValueData::IntervalMonthDayNanoValue(convert_i128_to_interval(v.to_i128()))
-                }
-            }),
+            Value::IntervalYearMonth(v) => Some(ValueData::IntervalYearMonthValue(v.to_i32())),
+            Value::IntervalDayTime(v) => Some(ValueData::IntervalDayTimeValue(v.to_i64())),
+            Value::IntervalMonthDayNano(v) => Some(ValueData::IntervalMonthDayNanoValue(
+                convert_month_day_nano_to_pb(v),
+            )),
             Value::Decimal128(v) => Some(ValueData::Decimal128Value(convert_to_pb_decimal128(v))),
             Value::List(_) | Value::Duration(_) => unreachable!(),
         },
@@ -1032,6 +1090,7 @@ pub fn value_to_grpc_value(value: Value) -> GrpcValue {
 mod tests {
     use std::sync::Arc;
 
+    use common_time::interval::IntervalUnit;
     use datatypes::types::{
         Int32Type, IntervalDayTimeType, IntervalMonthDayNanoType, IntervalYearMonthType,
         TimeMillisecondType, TimeSecondType, TimestampMillisecondType, TimestampSecondType,
@@ -1120,6 +1179,10 @@ mod tests {
         let values = values_with_capacity(ColumnDataType::Decimal128, 2);
         let values = values.decimal128_values;
         assert_eq!(2, values.capacity());
+
+        let values = values_with_capacity(ColumnDataType::Vector, 2);
+        let values = values.binary_values;
+        assert_eq!(2, values.capacity());
     }
 
     #[test]
@@ -1207,7 +1270,11 @@ mod tests {
         assert_eq!(
             ConcreteDataType::decimal128_datatype(10, 2),
             ColumnDataTypeWrapper::decimal128_datatype(10, 2).into()
-        )
+        );
+        assert_eq!(
+            ConcreteDataType::vector_datatype(3),
+            ColumnDataTypeWrapper::vector_datatype(3).into()
+        );
     }
 
     #[test]
@@ -1303,6 +1370,10 @@ mod tests {
                 .try_into()
                 .unwrap()
         );
+        assert_eq!(
+            ColumnDataTypeWrapper::vector_datatype(3),
+            ConcreteDataType::vector_datatype(3).try_into().unwrap()
+        );
 
         let result: Result<ColumnDataTypeWrapper> = ConcreteDataType::null_datatype().try_into();
         assert!(result.is_err());
@@ -1477,11 +1548,11 @@ mod tests {
 
     #[test]
     fn test_convert_i128_to_interval() {
-        let i128_val = 3000;
-        let interval = convert_i128_to_interval(i128_val);
+        let i128_val = 3;
+        let interval = convert_month_day_nano_to_pb(IntervalMonthDayNano::from_i128(i128_val));
         assert_eq!(interval.months, 0);
         assert_eq!(interval.days, 0);
-        assert_eq!(interval.nanoseconds, 3000);
+        assert_eq!(interval.nanoseconds, 3);
     }
 
     #[test]
@@ -1561,9 +1632,9 @@ mod tests {
             },
         );
         let expect = vec![
-            Value::Interval(Interval::from_year_month(1_i32)),
-            Value::Interval(Interval::from_year_month(2_i32)),
-            Value::Interval(Interval::from_year_month(3_i32)),
+            Value::IntervalYearMonth(IntervalYearMonth::new(1_i32)),
+            Value::IntervalYearMonth(IntervalYearMonth::new(2_i32)),
+            Value::IntervalYearMonth(IntervalYearMonth::new(3_i32)),
         ];
         assert_eq!(expect, actual);
 
@@ -1576,9 +1647,9 @@ mod tests {
             },
         );
         let expect = vec![
-            Value::Interval(Interval::from_i64(1_i64)),
-            Value::Interval(Interval::from_i64(2_i64)),
-            Value::Interval(Interval::from_i64(3_i64)),
+            Value::IntervalDayTime(IntervalDayTime::from_i64(1_i64)),
+            Value::IntervalDayTime(IntervalDayTime::from_i64(2_i64)),
+            Value::IntervalDayTime(IntervalDayTime::from_i64(3_i64)),
         ];
         assert_eq!(expect, actual);
 
@@ -1607,9 +1678,9 @@ mod tests {
             },
         );
         let expect = vec![
-            Value::Interval(Interval::from_month_day_nano(1, 2, 3)),
-            Value::Interval(Interval::from_month_day_nano(5, 6, 7)),
-            Value::Interval(Interval::from_month_day_nano(9, 10, 11)),
+            Value::IntervalMonthDayNano(IntervalMonthDayNano::new(1, 2, 3)),
+            Value::IntervalMonthDayNano(IntervalMonthDayNano::new(5, 6, 7)),
+            Value::IntervalMonthDayNano(IntervalMonthDayNano::new(9, 10, 11)),
         ];
         assert_eq!(expect, actual);
     }

src/api/src/region.rs

@@ -21,14 +21,14 @@ use greptime_proto::v1::region::RegionResponse as RegionResponseV1;
 #[derive(Debug)]
 pub struct RegionResponse {
     pub affected_rows: AffectedRows,
-    pub extension: HashMap<String, Vec<u8>>,
+    pub extensions: HashMap<String, Vec<u8>>,
 }
 
 impl RegionResponse {
     pub fn from_region_response(region_response: RegionResponseV1) -> Self {
         Self {
             affected_rows: region_response.affected_rows as _,
-            extension: region_response.extension,
+            extensions: region_response.extensions,
         }
     }
 
@@ -36,7 +36,7 @@ impl RegionResponse {
     pub fn new(affected_rows: AffectedRows) -> Self {
         Self {
             affected_rows,
-            extension: Default::default(),
+            extensions: Default::default(),
         }
     }
 }

src/api/src/v1/column_def.rs

@@ -15,8 +15,10 @@
 use std::collections::HashMap;
 
 use datatypes::schema::{
-    ColumnDefaultConstraint, ColumnSchema, FulltextOptions, COMMENT_KEY, FULLTEXT_KEY,
+    ColumnDefaultConstraint, ColumnSchema, FulltextAnalyzer, FulltextOptions, COMMENT_KEY,
+    FULLTEXT_KEY, INVERTED_INDEX_KEY,
 };
+use greptime_proto::v1::Analyzer;
 use snafu::ResultExt;
 
 use crate::error::{self, Result};
@@ -25,6 +27,8 @@ use crate::v1::{ColumnDef, ColumnOptions, SemanticType};
 
 /// Key used to store fulltext options in gRPC column options.
 const FULLTEXT_GRPC_KEY: &str = "fulltext";
+/// Key used to store inverted index options in gRPC column options.
+const INVERTED_INDEX_GRPC_KEY: &str = "inverted_index";
 
 /// Tries to construct a `ColumnSchema` from the given  `ColumnDef`.
 pub fn try_as_column_schema(column_def: &ColumnDef) -> Result<ColumnSchema> {
@@ -49,10 +53,13 @@ pub fn try_as_column_schema(column_def: &ColumnDef) -> Result<ColumnSchema> {
     if !column_def.comment.is_empty() {
         metadata.insert(COMMENT_KEY.to_string(), column_def.comment.clone());
     }
-    if let Some(options) = column_def.options.as_ref()
-        && let Some(fulltext) = options.options.get(FULLTEXT_GRPC_KEY)
-    {
-        metadata.insert(FULLTEXT_KEY.to_string(), fulltext.to_string());
+    if let Some(options) = column_def.options.as_ref() {
+        if let Some(fulltext) = options.options.get(FULLTEXT_GRPC_KEY) {
+            metadata.insert(FULLTEXT_KEY.to_string(), fulltext.clone());
+        }
+        if let Some(inverted_index) = options.options.get(INVERTED_INDEX_GRPC_KEY) {
+            metadata.insert(INVERTED_INDEX_KEY.to_string(), inverted_index.clone());
+        }
     }
 
     ColumnSchema::new(&column_def.name, data_type.into(), column_def.is_nullable)
@@ -70,7 +77,12 @@ pub fn options_from_column_schema(column_schema: &ColumnSchema) -> Option<Column
     if let Some(fulltext) = column_schema.metadata().get(FULLTEXT_KEY) {
         options
             .options
-            .insert(FULLTEXT_GRPC_KEY.to_string(), fulltext.to_string());
+            .insert(FULLTEXT_GRPC_KEY.to_string(), fulltext.clone());
+    }
+    if let Some(inverted_index) = column_schema.metadata().get(INVERTED_INDEX_KEY) {
+        options
+            .options
+            .insert(INVERTED_INDEX_GRPC_KEY.to_string(), inverted_index.clone());
     }
 
     (!options.options.is_empty()).then_some(options)
@@ -93,6 +105,14 @@ pub fn options_from_fulltext(fulltext: &FulltextOptions) -> Result<Option<Column
     Ok((!options.options.is_empty()).then_some(options))
 }
 
+/// Tries to construct a `FulltextAnalyzer` from the given analyzer.
+pub fn as_fulltext_option(analyzer: Analyzer) -> FulltextAnalyzer {
+    match analyzer {
+        Analyzer::English => FulltextAnalyzer::English,
+        Analyzer::Chinese => FulltextAnalyzer::Chinese,
+    }
+}
+
 #[cfg(test)]
 mod tests {
 
@@ -115,10 +135,13 @@ mod tests {
             comment: "test_comment".to_string(),
             datatype_extension: None,
             options: Some(ColumnOptions {
-                options: HashMap::from([(
-                    FULLTEXT_GRPC_KEY.to_string(),
-                    "{\"enable\":true}".to_string(),
-                )]),
+                options: HashMap::from([
+                    (
+                        FULLTEXT_GRPC_KEY.to_string(),
+                        "{\"enable\":true}".to_string(),
+                    ),
+                    (INVERTED_INDEX_GRPC_KEY.to_string(), "true".to_string()),
+                ]),
             }),
         };
 
@@ -139,6 +162,7 @@ mod tests {
                 ..Default::default()
             }
         );
+        assert!(schema.is_inverted_indexed());
     }
 
     #[test]
@@ -153,12 +177,17 @@ mod tests {
                 analyzer: FulltextAnalyzer::English,
                 case_sensitive: false,
             })
-            .unwrap();
+            .unwrap()
+            .set_inverted_index(true);
         let options = options_from_column_schema(&schema).unwrap();
         assert_eq!(
             options.options.get(FULLTEXT_GRPC_KEY).unwrap(),
             "{\"enable\":true,\"analyzer\":\"English\",\"case-sensitive\":false}"
         );
+        assert_eq!(
+            options.options.get(INVERTED_INDEX_GRPC_KEY).unwrap(),
+            "true"
+        );
     }
 
     #[test]

src/auth/src/common.rs

@@ -75,6 +75,16 @@ pub enum Password<'a> {
     PgMD5(HashedPassword<'a>, Salt<'a>),
 }
 
+impl Password<'_> {
+    pub fn r#type(&self) -> &str {
+        match self {
+            Password::PlainText(_) => "plain_text",
+            Password::MysqlNativePassword(_, _) => "mysql_native_password",
+            Password::PgMD5(_, _) => "pg_md5",
+        }
+    }
+}
+
 pub fn auth_mysql(
     auth_data: HashedPassword,
     salt: Salt,

src/auth/src/error.rs

@@ -89,7 +89,7 @@ impl ErrorExt for Error {
             Error::FileWatch { .. } => StatusCode::InvalidArguments,
             Error::InternalState { .. } => StatusCode::Unexpected,
             Error::Io { .. } => StatusCode::StorageUnavailable,
-            Error::AuthBackend { .. } => StatusCode::Internal,
+            Error::AuthBackend { source, .. } => source.status_code(),
 
             Error::UserNotFound { .. } => StatusCode::UserNotFound,
             Error::UnsupportedPasswordType { .. } => StatusCode::UnsupportedPasswordType,

src/auth/src/tests.rs

@@ -13,9 +13,11 @@
 // limitations under the License.
 
 use common_base::secrets::ExposeSecret;
+use common_error::ext::BoxedError;
+use snafu::{OptionExt, ResultExt};
 
 use crate::error::{
-    AccessDeniedSnafu, Result, UnsupportedPasswordTypeSnafu, UserNotFoundSnafu,
+    AccessDeniedSnafu, AuthBackendSnafu, Result, UnsupportedPasswordTypeSnafu, UserNotFoundSnafu,
     UserPasswordMismatchSnafu,
 };
 use crate::user_info::DefaultUserInfo;
@@ -49,6 +51,19 @@ impl MockUserProvider {
         info.schema.clone_into(&mut self.schema);
         info.username.clone_into(&mut self.username);
     }
+
+    // this is a deliberate function to ref AuthBackendSnafu
+    // so that it won't get deleted in the future
+    pub fn ref_auth_backend_snafu(&self) -> Result<()> {
+        let none_option = None;
+
+        none_option
+            .context(UserNotFoundSnafu {
+                username: "no_user".to_string(),
+            })
+            .map_err(BoxedError::new)
+            .context(AuthBackendSnafu)
+    }
 }
 
 #[async_trait::async_trait]

src/auth/src/user_provider.rs

@@ -57,6 +57,11 @@ pub trait UserProvider: Send + Sync {
         self.authorize(catalog, schema, &user_info).await?;
         Ok(user_info)
     }
+
+    /// Returns whether this user provider implementation is backed by an external system.
+    fn external(&self) -> bool {
+        false
+    }
 }
 
 fn load_credential_from_file(filepath: &str) -> Result<Option<HashMap<String, Vec<u8>>>> {

src/auth/src/user_provider/static_user_provider.rs

@@ -33,7 +33,7 @@ impl StaticUserProvider {
             value: value.to_string(),
             msg: "StaticUserProviderOption must be in format `<option>:<value>`",
         })?;
-        return match mode {
+        match mode {
             "file" => {
                 let users = load_credential_from_file(content)?
                     .context(InvalidConfigSnafu {
@@ -58,7 +58,7 @@ impl StaticUserProvider {
                 msg: "StaticUserProviderOption must be in format `file:<path>` or `cmd:<values>`",
             }
                 .fail(),
-        };
+        }
     }
 }
 

src/auth/tests/mod.rs

@@ -18,6 +18,7 @@ use std::sync::Arc;
 
 use api::v1::greptime_request::Request;
 use auth::error::Error::InternalState;
+use auth::error::InternalStateSnafu;
 use auth::{PermissionChecker, PermissionCheckerRef, PermissionReq, PermissionResp, UserInfoRef};
 use sql::statements::show::{ShowDatabases, ShowKind};
 use sql::statements::statement::Statement;
@@ -33,9 +34,10 @@ impl PermissionChecker for DummyPermissionChecker {
         match req {
             PermissionReq::GrpcRequest(_) => Ok(PermissionResp::Allow),
             PermissionReq::SqlStatement(_) => Ok(PermissionResp::Reject),
-            _ => Err(InternalState {
+            _ => InternalStateSnafu {
                 msg: "testing".to_string(),
-            }),
+            }
+            .fail(),
         }
     }
 }

src/cache/src/lib.rs

@@ -19,9 +19,9 @@ use std::time::Duration;
 
 use catalog::kvbackend::new_table_cache;
 use common_meta::cache::{
-    new_table_flownode_set_cache, new_table_info_cache, new_table_name_cache,
-    new_table_route_cache, new_view_info_cache, CacheRegistry, CacheRegistryBuilder,
-    LayeredCacheRegistryBuilder,
+    new_schema_cache, new_table_flownode_set_cache, new_table_info_cache, new_table_name_cache,
+    new_table_route_cache, new_table_schema_cache, new_view_info_cache, CacheRegistry,
+    CacheRegistryBuilder, LayeredCacheRegistryBuilder,
 };
 use common_meta::kv_backend::KvBackendRef;
 use moka::future::CacheBuilder;
@@ -37,9 +37,47 @@ pub const TABLE_INFO_CACHE_NAME: &str = "table_info_cache";
 pub const VIEW_INFO_CACHE_NAME: &str = "view_info_cache";
 pub const TABLE_NAME_CACHE_NAME: &str = "table_name_cache";
 pub const TABLE_CACHE_NAME: &str = "table_cache";
+pub const SCHEMA_CACHE_NAME: &str = "schema_cache";
+pub const TABLE_SCHEMA_NAME_CACHE_NAME: &str = "table_schema_name_cache";
 pub const TABLE_FLOWNODE_SET_CACHE_NAME: &str = "table_flownode_set_cache";
 pub const TABLE_ROUTE_CACHE_NAME: &str = "table_route_cache";
 
+/// Builds cache registry for datanode, including:
+/// - Schema cache.
+/// - Table id to schema name cache.
+pub fn build_datanode_cache_registry(kv_backend: KvBackendRef) -> CacheRegistry {
+    // Builds table id schema name cache that never expires.
+    let cache = CacheBuilder::new(DEFAULT_CACHE_MAX_CAPACITY).build();
+    let table_id_schema_cache = Arc::new(new_table_schema_cache(
+        TABLE_SCHEMA_NAME_CACHE_NAME.to_string(),
+        cache,
+        kv_backend.clone(),
+    ));
+
+    // Builds schema cache
+    let cache = CacheBuilder::new(DEFAULT_CACHE_MAX_CAPACITY)
+        .time_to_live(DEFAULT_CACHE_TTL)
+        .time_to_idle(DEFAULT_CACHE_TTI)
+        .build();
+    let schema_cache = Arc::new(new_schema_cache(
+        SCHEMA_CACHE_NAME.to_string(),
+        cache,
+        kv_backend.clone(),
+    ));
+
+    CacheRegistryBuilder::default()
+        .add_cache(table_id_schema_cache)
+        .add_cache(schema_cache)
+        .build()
+}
+
+/// Builds cache registry for frontend and datanode, including:
+/// - Table info cache
+/// - Table name cache
+/// - Table route cache
+/// - Table flow node cache
+/// - View cache
+/// - Schema cache
 pub fn build_fundamental_cache_registry(kv_backend: KvBackendRef) -> CacheRegistry {
     // Builds table info cache
     let cache = CacheBuilder::new(DEFAULT_CACHE_MAX_CAPACITY)
@@ -95,12 +133,30 @@ pub fn build_fundamental_cache_registry(kv_backend: KvBackendRef) -> CacheRegist
         kv_backend.clone(),
     ));
 
+    // Builds schema cache
+    let cache = CacheBuilder::new(DEFAULT_CACHE_MAX_CAPACITY)
+        .time_to_live(DEFAULT_CACHE_TTL)
+        .time_to_idle(DEFAULT_CACHE_TTI)
+        .build();
+    let schema_cache = Arc::new(new_schema_cache(
+        SCHEMA_CACHE_NAME.to_string(),
+        cache,
+        kv_backend.clone(),
+    ));
+
+    let table_id_schema_cache = Arc::new(new_table_schema_cache(
+        TABLE_SCHEMA_NAME_CACHE_NAME.to_string(),
+        CacheBuilder::new(DEFAULT_CACHE_MAX_CAPACITY).build(),
+        kv_backend,
+    ));
     CacheRegistryBuilder::default()
         .add_cache(table_info_cache)
         .add_cache(table_name_cache)
         .add_cache(table_route_cache)
         .add_cache(view_info_cache)
         .add_cache(table_flownode_set_cache)
+        .add_cache(schema_cache)
+        .add_cache(table_id_schema_cache)
         .build()
 }
 

src/catalog/Cargo.toml

@@ -22,8 +22,10 @@ common-config.workspace = true
 common-error.workspace = true
 common-macro.workspace = true
 common-meta.workspace = true
+common-procedure.workspace = true
 common-query.workspace = true
 common-recordbatch.workspace = true
+common-runtime.workspace = true
 common-telemetry.workspace = true
 common-time.workspace = true
 common-version.workspace = true
@@ -48,6 +50,7 @@ sql.workspace = true
 store-api.workspace = true
 table.workspace = true
 tokio.workspace = true
+tokio-stream = "0.1"
 
 [dev-dependencies]
 cache.workspace = true

src/catalog/src/error.rs

@@ -50,13 +50,20 @@ pub enum Error {
         source: BoxedError,
     },
 
-    #[snafu(display("Failed to list nodes in cluster: {source}"))]
+    #[snafu(display("Failed to list nodes in cluster"))]
     ListNodes {
         #[snafu(implicit)]
         location: Location,
         source: BoxedError,
     },
 
+    #[snafu(display("Failed to region stats in cluster"))]
+    ListRegionStats {
+        #[snafu(implicit)]
+        location: Location,
+        source: BoxedError,
+    },
+
     #[snafu(display("Failed to list flows in catalog {catalog}"))]
     ListFlows {
         #[snafu(implicit)]
@@ -82,6 +89,32 @@ pub enum Error {
         location: Location,
     },
 
+    #[snafu(display("Failed to get information extension client"))]
+    GetInformationExtension {
+        #[snafu(implicit)]
+        location: Location,
+    },
+
+    #[snafu(display("Failed to list procedures"))]
+    ListProcedures {
+        #[snafu(implicit)]
+        location: Location,
+        source: BoxedError,
+    },
+
+    #[snafu(display("Procedure id not found"))]
+    ProcedureIdNotFound {
+        #[snafu(implicit)]
+        location: Location,
+    },
+
+    #[snafu(display("convert proto data error"))]
+    ConvertProtoData {
+        #[snafu(implicit)]
+        location: Location,
+        source: BoxedError,
+    },
+
     #[snafu(display("Failed to re-compile script due to internal error"))]
     CompileScriptInternal {
         #[snafu(implicit)]
@@ -97,13 +130,6 @@ pub enum Error {
         source: table::error::Error,
     },
 
-    #[snafu(display("System catalog is not valid: {}", msg))]
-    SystemCatalog {
-        msg: String,
-        #[snafu(implicit)]
-        location: Location,
-    },
-
     #[snafu(display("Cannot find catalog by name: {}", catalog_name))]
     CatalogNotFound {
         catalog_name: String,
@@ -152,6 +178,12 @@ pub enum Error {
         location: Location,
     },
 
+    #[snafu(display("Partition manager not found, it's not expected."))]
+    PartitionManagerNotFound {
+        #[snafu(implicit)]
+        location: Location,
+    },
+
     #[snafu(display("Failed to find table partitions"))]
     FindPartitions { source: partition::error::Error },
 
@@ -186,13 +218,6 @@ pub enum Error {
         source: common_query::error::Error,
     },
 
-    #[snafu(display("Failed to perform metasrv operation"))]
-    Metasrv {
-        #[snafu(implicit)]
-        location: Location,
-        source: meta_client::error::Error,
-    },
-
     #[snafu(display("Invalid table info in catalog"))]
     InvalidTableInfoInCatalog {
         #[snafu(implicit)]
@@ -280,7 +305,10 @@ impl ErrorExt for Error {
             | Error::FindRegionRoutes { .. }
             | Error::CacheNotFound { .. }
             | Error::CastManager { .. }
-            | Error::Json { .. } => StatusCode::Unexpected,
+            | Error::Json { .. }
+            | Error::GetInformationExtension { .. }
+            | Error::PartitionManagerNotFound { .. }
+            | Error::ProcedureIdNotFound { .. } => StatusCode::Unexpected,
 
             Error::ViewPlanColumnsChanged { .. } => StatusCode::InvalidArguments,
 
@@ -288,8 +316,6 @@ impl ErrorExt for Error {
 
             Error::FlowInfoNotFound { .. } => StatusCode::FlowNotFound,
 
-            Error::SystemCatalog { .. } => StatusCode::StorageUnavailable,
-
             Error::UpgradeWeakCatalogManagerRef { .. } => StatusCode::Internal,
 
             Error::CreateRecordBatch { source, .. } => source.status_code(),
@@ -299,11 +325,13 @@ impl ErrorExt for Error {
             | Error::ListNodes { source, .. }
             | Error::ListSchemas { source, .. }
             | Error::ListTables { source, .. }
-            | Error::ListFlows { source, .. } => source.status_code(),
+            | Error::ListFlows { source, .. }
+            | Error::ListProcedures { source, .. }
+            | Error::ListRegionStats { source, .. }
+            | Error::ConvertProtoData { source, .. } => source.status_code(),
 
             Error::CreateTable { source, .. } => source.status_code(),
 
-            Error::Metasrv { source, .. } => source.status_code(),
             Error::DecodePlan { source, .. } => source.status_code(),
             Error::InvalidTableInfoInCatalog { source, .. } => source.status_code(),
 
@@ -338,27 +366,6 @@ mod tests {
 
     use super::*;
 
-    #[test]
-    pub fn test_error_status_code() {
-        assert_eq!(
-            StatusCode::TableAlreadyExists,
-            Error::TableExists {
-                table: "some_table".to_string(),
-                location: Location::generate(),
-            }
-            .status_code()
-        );
-
-        assert_eq!(
-            StatusCode::StorageUnavailable,
-            Error::SystemCatalog {
-                msg: String::default(),
-                location: Location::generate(),
-            }
-            .status_code()
-        );
-    }
-
     #[test]
     pub fn test_errors_to_datafusion_error() {
         let e: DataFusionError = Error::TableExists {

src/catalog/src/information_extension.rs

@@ -0,0 +1,92 @@
+// Copyright 2023 Greptime Team
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//     http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+use api::v1::meta::ProcedureStatus;
+use common_error::ext::BoxedError;
+use common_meta::cluster::{ClusterInfo, NodeInfo};
+use common_meta::datanode::RegionStat;
+use common_meta::ddl::{ExecutorContext, ProcedureExecutor};
+use common_meta::rpc::procedure;
+use common_procedure::{ProcedureInfo, ProcedureState};
+use meta_client::MetaClientRef;
+use snafu::ResultExt;
+
+use crate::error;
+use crate::information_schema::InformationExtension;
+
+pub struct DistributedInformationExtension {
+    meta_client: MetaClientRef,
+}
+
+impl DistributedInformationExtension {
+    pub fn new(meta_client: MetaClientRef) -> Self {
+        Self { meta_client }
+    }
+}
+
+#[async_trait::async_trait]
+impl InformationExtension for DistributedInformationExtension {
+    type Error = crate::error::Error;
+
+    async fn nodes(&self) -> std::result::Result<Vec<NodeInfo>, Self::Error> {
+        self.meta_client
+            .list_nodes(None)
+            .await
+            .map_err(BoxedError::new)
+            .context(error::ListNodesSnafu)
+    }
+
+    async fn procedures(&self) -> std::result::Result<Vec<(String, ProcedureInfo)>, Self::Error> {
+        let procedures = self
+            .meta_client
+            .list_procedures(&ExecutorContext::default())
+            .await
+            .map_err(BoxedError::new)
+            .context(error::ListProceduresSnafu)?
+            .procedures;
+        let mut result = Vec::with_capacity(procedures.len());
+        for procedure in procedures {
+            let pid = match procedure.id {
+                Some(pid) => pid,
+                None => return error::ProcedureIdNotFoundSnafu {}.fail(),
+            };
+            let pid = procedure::pb_pid_to_pid(&pid)
+                .map_err(BoxedError::new)
+                .context(error::ConvertProtoDataSnafu)?;
+            let status = ProcedureStatus::try_from(procedure.status)
+                .map(|v| v.as_str_name())
+                .unwrap_or("Unknown")
+                .to_string();
+            let procedure_info = ProcedureInfo {
+                id: pid,
+                type_name: procedure.type_name,
+                start_time_ms: procedure.start_time_ms,
+                end_time_ms: procedure.end_time_ms,
+                state: ProcedureState::Running,
+                lock_keys: procedure.lock_keys,
+            };
+            result.push((status, procedure_info));
+        }
+
+        Ok(result)
+    }
+
+    async fn region_stats(&self) -> std::result::Result<Vec<RegionStat>, Self::Error> {
+        self.meta_client
+            .list_region_stats()
+            .await
+            .map_err(BoxedError::new)
+            .context(error::ListRegionStatsSnafu)
+    }
+}

src/catalog/src/kvbackend.rs

@@ -12,7 +12,7 @@
 // See the License for the specific language governing permissions and
 // limitations under the License.
 
-pub use client::{CachedMetaKvBackend, CachedMetaKvBackendBuilder, MetaKvBackend};
+pub use client::{CachedKvBackend, CachedKvBackendBuilder, MetaKvBackend};
 
 mod client;
 mod manager;

src/catalog/src/kvbackend/client.rs

@@ -20,8 +20,9 @@ use std::time::Duration;
 
 use common_error::ext::BoxedError;
 use common_meta::cache_invalidator::KvCacheInvalidator;
-use common_meta::error::Error::{CacheNotGet, GetKvCache};
-use common_meta::error::{CacheNotGetSnafu, Error, ExternalSnafu, Result};
+use common_meta::error::Error::CacheNotGet;
+use common_meta::error::{CacheNotGetSnafu, Error, ExternalSnafu, GetKvCacheSnafu, Result};
+use common_meta::kv_backend::txn::{Txn, TxnResponse};
 use common_meta::kv_backend::{KvBackend, KvBackendRef, TxnService};
 use common_meta::rpc::store::{
     BatchDeleteRequest, BatchDeleteResponse, BatchGetRequest, BatchGetResponse, BatchPutRequest,
@@ -42,20 +43,20 @@ const DEFAULT_CACHE_MAX_CAPACITY: u64 = 10000;
 const DEFAULT_CACHE_TTL: Duration = Duration::from_secs(10 * 60);
 const DEFAULT_CACHE_TTI: Duration = Duration::from_secs(5 * 60);
 
-pub struct CachedMetaKvBackendBuilder {
+pub struct CachedKvBackendBuilder {
     cache_max_capacity: Option<u64>,
     cache_ttl: Option<Duration>,
     cache_tti: Option<Duration>,
-    meta_client: Arc<MetaClient>,
+    inner: KvBackendRef,
 }
 
-impl CachedMetaKvBackendBuilder {
-    pub fn new(meta_client: Arc<MetaClient>) -> Self {
+impl CachedKvBackendBuilder {
+    pub fn new(inner: KvBackendRef) -> Self {
         Self {
             cache_max_capacity: None,
             cache_ttl: None,
             cache_tti: None,
-            meta_client,
+            inner,
         }
     }
 
@@ -74,7 +75,7 @@ impl CachedMetaKvBackendBuilder {
         self
     }
 
-    pub fn build(self) -> CachedMetaKvBackend {
+    pub fn build(self) -> CachedKvBackend {
         let cache_max_capacity = self
             .cache_max_capacity
             .unwrap_or(DEFAULT_CACHE_MAX_CAPACITY);
@@ -85,14 +86,11 @@ impl CachedMetaKvBackendBuilder {
             .time_to_live(cache_ttl)
             .time_to_idle(cache_tti)
             .build();
-
-        let kv_backend = Arc::new(MetaKvBackend {
-            client: self.meta_client,
-        });
+        let kv_backend = self.inner;
         let name = format!("CachedKvBackend({})", kv_backend.name());
         let version = AtomicUsize::new(0);
 
-        CachedMetaKvBackend {
+        CachedKvBackend {
             kv_backend,
             cache,
             name,
@@ -112,19 +110,29 @@ pub type CacheBackend = Cache<Vec<u8>, KeyValue>;
 /// Therefore, it is recommended to use CachedMetaKvBackend to only read metadata related
 /// information. Note: If you read other information, you may read expired data, which depends on
 /// TTL and TTI for cache.
-pub struct CachedMetaKvBackend {
+pub struct CachedKvBackend {
     kv_backend: KvBackendRef,
     cache: CacheBackend,
     name: String,
     version: AtomicUsize,
 }
 
-impl TxnService for CachedMetaKvBackend {
+#[async_trait::async_trait]
+impl TxnService for CachedKvBackend {
     type Error = Error;
+
+    async fn txn(&self, txn: Txn) -> std::result::Result<TxnResponse, Self::Error> {
+        // TODO(hl): txn of CachedKvBackend simply pass through to inner backend without invalidating caches.
+        self.kv_backend.txn(txn).await
+    }
+
+    fn max_txn_ops(&self) -> usize {
+        self.kv_backend.max_txn_ops()
+    }
 }
 
 #[async_trait::async_trait]
-impl KvBackend for CachedMetaKvBackend {
+impl KvBackend for CachedKvBackend {
     fn name(&self) -> &str {
         &self.name
     }
@@ -282,8 +290,11 @@ impl KvBackend for CachedMetaKvBackend {
                 _ => Err(e),
             },
         }
-        .map_err(|e| GetKvCache {
-            err_msg: e.to_string(),
+        .map_err(|e| {
+            GetKvCacheSnafu {
+                err_msg: e.to_string(),
+            }
+            .build()
         });
 
         // "cache.invalidate_key" and "cache.try_get_with_by_ref" are not mutually exclusive. So we need
@@ -302,7 +313,7 @@ impl KvBackend for CachedMetaKvBackend {
 }
 
 #[async_trait::async_trait]
-impl KvCacheInvalidator for CachedMetaKvBackend {
+impl KvCacheInvalidator for CachedKvBackend {
     async fn invalidate_key(&self, key: &[u8]) {
         self.create_new_version();
         self.cache.invalidate(key).await;
@@ -310,7 +321,7 @@ impl KvCacheInvalidator for CachedMetaKvBackend {
     }
 }
 
-impl CachedMetaKvBackend {
+impl CachedKvBackend {
     // only for test
     #[cfg(test)]
     fn wrap(kv_backend: KvBackendRef) -> Self {
@@ -463,7 +474,7 @@ mod tests {
     use common_meta::rpc::KeyValue;
     use dashmap::DashMap;
 
-    use super::CachedMetaKvBackend;
+    use super::CachedKvBackend;
 
     #[derive(Default)]
     pub struct SimpleKvBackend {
@@ -537,7 +548,7 @@ mod tests {
     async fn test_cached_kv_backend() {
         let simple_kv = Arc::new(SimpleKvBackend::default());
         let get_execute_times = simple_kv.get_execute_times.clone();
-        let cached_kv = CachedMetaKvBackend::wrap(simple_kv);
+        let cached_kv = CachedKvBackend::wrap(simple_kv);
 
         add_some_vals(&cached_kv).await;
 

src/catalog/src/kvbackend/manager.rs

@@ -21,7 +21,6 @@ use common_catalog::consts::{
     DEFAULT_CATALOG_NAME, DEFAULT_SCHEMA_NAME, INFORMATION_SCHEMA_NAME, NUMBERS_TABLE_ID,
     PG_CATALOG_NAME,
 };
-use common_config::Mode;
 use common_error::ext::BoxedError;
 use common_meta::cache::{LayeredCacheRegistryRef, ViewInfoCacheRef};
 use common_meta::key::catalog_name::CatalogNameKey;
@@ -31,22 +30,25 @@ use common_meta::key::table_info::TableInfoValue;
 use common_meta::key::table_name::TableNameKey;
 use common_meta::key::{TableMetadataManager, TableMetadataManagerRef};
 use common_meta::kv_backend::KvBackendRef;
+use common_procedure::ProcedureManagerRef;
 use futures_util::stream::BoxStream;
 use futures_util::{StreamExt, TryStreamExt};
-use meta_client::client::MetaClient;
 use moka::sync::Cache;
 use partition::manager::{PartitionRuleManager, PartitionRuleManagerRef};
+use session::context::{Channel, QueryContext};
 use snafu::prelude::*;
 use table::dist_table::DistTable;
 use table::table::numbers::{NumbersTable, NUMBERS_TABLE_NAME};
 use table::table_name::TableName;
 use table::TableRef;
+use tokio::sync::Semaphore;
+use tokio_stream::wrappers::ReceiverStream;
 
 use crate::error::{
     CacheNotFoundSnafu, GetTableCacheSnafu, InvalidTableInfoInCatalogSnafu, ListCatalogsSnafu,
     ListSchemasSnafu, ListTablesSnafu, Result, TableMetadataManagerSnafu,
 };
-use crate::information_schema::InformationSchemaProvider;
+use crate::information_schema::{InformationExtensionRef, InformationSchemaProvider};
 use crate::kvbackend::TableCacheRef;
 use crate::system_schema::pg_catalog::PGCatalogProvider;
 use crate::system_schema::SystemSchemaProvider;
@@ -59,27 +61,31 @@ use crate::CatalogManager;
 /// comes from `SystemCatalog`, which is static and read-only.
 #[derive(Clone)]
 pub struct KvBackendCatalogManager {
-    mode: Mode,
-    meta_client: Option<Arc<MetaClient>>,
+    /// Provides the extension methods for the `information_schema` tables
+    information_extension: InformationExtensionRef,
+    /// Manages partition rules.
     partition_manager: PartitionRuleManagerRef,
+    /// Manages table metadata.
     table_metadata_manager: TableMetadataManagerRef,
     /// A sub-CatalogManager that handles system tables
     system_catalog: SystemCatalog,
+    /// Cache registry for all caches.
     cache_registry: LayeredCacheRegistryRef,
+    /// Only available in `Standalone` mode.
+    procedure_manager: Option<ProcedureManagerRef>,
 }
 
 const CATALOG_CACHE_MAX_CAPACITY: u64 = 128;
 
 impl KvBackendCatalogManager {
     pub fn new(
-        mode: Mode,
-        meta_client: Option<Arc<MetaClient>>,
+        information_extension: InformationExtensionRef,
         backend: KvBackendRef,
         cache_registry: LayeredCacheRegistryRef,
+        procedure_manager: Option<ProcedureManagerRef>,
     ) -> Arc<Self> {
         Arc::new_cyclic(|me| Self {
-            mode,
-            meta_client,
+            information_extension,
             partition_manager: Arc::new(PartitionRuleManager::new(
                 backend.clone(),
                 cache_registry
@@ -103,23 +109,19 @@ impl KvBackendCatalogManager {
                 backend,
             },
             cache_registry,
+            procedure_manager,
         })
     }
 
-    /// Returns the server running mode.
-    pub fn running_mode(&self) -> &Mode {
-        &self.mode
-    }
-
     pub fn view_info_cache(&self) -> Result<ViewInfoCacheRef> {
         self.cache_registry.get().context(CacheNotFoundSnafu {
             name: "view_info_cache",
         })
     }
 
-    /// Returns the `[MetaClient]`.
-    pub fn meta_client(&self) -> Option<Arc<MetaClient>> {
-        self.meta_client.clone()
+    /// Returns the [`InformationExtension`].
+    pub fn information_extension(&self) -> InformationExtensionRef {
+        self.information_extension.clone()
     }
 
     pub fn partition_manager(&self) -> PartitionRuleManagerRef {
@@ -129,6 +131,10 @@ impl KvBackendCatalogManager {
     pub fn table_metadata_manager_ref(&self) -> &TableMetadataManagerRef {
         &self.table_metadata_manager
     }
+
+    pub fn procedure_manager(&self) -> Option<ProcedureManagerRef> {
+        self.procedure_manager.clone()
+    }
 }
 
 #[async_trait::async_trait]
@@ -152,7 +158,11 @@ impl CatalogManager for KvBackendCatalogManager {
         Ok(keys)
     }
 
-    async fn schema_names(&self, catalog: &str) -> Result<Vec<String>> {
+    async fn schema_names(
+        &self,
+        catalog: &str,
+        query_ctx: Option<&QueryContext>,
+    ) -> Result<Vec<String>> {
         let stream = self
             .table_metadata_manager
             .schema_manager()
@@ -163,27 +173,29 @@ impl CatalogManager for KvBackendCatalogManager {
             .map_err(BoxedError::new)
             .context(ListSchemasSnafu { catalog })?;
 
-        keys.extend(self.system_catalog.schema_names());
+        keys.extend(self.system_catalog.schema_names(query_ctx));
 
         Ok(keys.into_iter().collect())
     }
 
-    async fn table_names(&self, catalog: &str, schema: &str) -> Result<Vec<String>> {
-        let stream = self
+    async fn table_names(
+        &self,
+        catalog: &str,
+        schema: &str,
+        query_ctx: Option<&QueryContext>,
+    ) -> Result<Vec<String>> {
+        let mut tables = self
             .table_metadata_manager
             .table_name_manager()
-            .tables(catalog, schema);
-        let mut tables = stream
+            .tables(catalog, schema)
+            .map_ok(|(table_name, _)| table_name)
             .try_collect::<Vec<_>>()
             .await
             .map_err(BoxedError::new)
-            .context(ListTablesSnafu { catalog, schema })?
-            .into_iter()
-            .map(|(k, _)| k)
-            .collect::<Vec<_>>();
-        tables.extend_from_slice(&self.system_catalog.table_names(schema));
+            .context(ListTablesSnafu { catalog, schema })?;
 
-        Ok(tables.into_iter().collect())
+        tables.extend(self.system_catalog.table_names(schema, query_ctx));
+        Ok(tables)
     }
 
     async fn catalog_exists(&self, catalog: &str) -> Result<bool> {
@@ -194,8 +206,13 @@ impl CatalogManager for KvBackendCatalogManager {
             .context(TableMetadataManagerSnafu)
     }
 
-    async fn schema_exists(&self, catalog: &str, schema: &str) -> Result<bool> {
-        if self.system_catalog.schema_exists(schema) {
+    async fn schema_exists(
+        &self,
+        catalog: &str,
+        schema: &str,
+        query_ctx: Option<&QueryContext>,
+    ) -> Result<bool> {
+        if self.system_catalog.schema_exists(schema, query_ctx) {
             return Ok(true);
         }
 
@@ -206,8 +223,14 @@ impl CatalogManager for KvBackendCatalogManager {
             .context(TableMetadataManagerSnafu)
     }
 
-    async fn table_exists(&self, catalog: &str, schema: &str, table: &str) -> Result<bool> {
-        if self.system_catalog.table_exists(schema, table) {
+    async fn table_exists(
+        &self,
+        catalog: &str,
+        schema: &str,
+        table: &str,
+        query_ctx: Option<&QueryContext>,
+    ) -> Result<bool> {
+        if self.system_catalog.table_exists(schema, table, query_ctx) {
             return Ok(true);
         }
 
@@ -225,10 +248,12 @@ impl CatalogManager for KvBackendCatalogManager {
         catalog_name: &str,
         schema_name: &str,
         table_name: &str,
+        query_ctx: Option<&QueryContext>,
     ) -> Result<Option<TableRef>> {
-        if let Some(table) = self
-            .system_catalog
-            .table(catalog_name, schema_name, table_name)
+        let channel = query_ctx.map_or(Channel::Unknown, |ctx| ctx.channel());
+        if let Some(table) =
+            self.system_catalog
+                .table(catalog_name, schema_name, table_name, query_ctx)
         {
             return Ok(Some(table));
         }
@@ -236,58 +261,112 @@ impl CatalogManager for KvBackendCatalogManager {
         let table_cache: TableCacheRef = self.cache_registry.get().context(CacheNotFoundSnafu {
             name: "table_cache",
         })?;
-
-        table_cache
+        if let Some(table) = table_cache
             .get_by_ref(&TableName {
                 catalog_name: catalog_name.to_string(),
                 schema_name: schema_name.to_string(),
                 table_name: table_name.to_string(),
             })
             .await
-            .context(GetTableCacheSnafu)
+            .context(GetTableCacheSnafu)?
+        {
+            return Ok(Some(table));
+        }
+
+        if channel == Channel::Postgres {
+            // falldown to pg_catalog
+            if let Some(table) =
+                self.system_catalog
+                    .table(catalog_name, PG_CATALOG_NAME, table_name, query_ctx)
+            {
+                return Ok(Some(table));
+            }
+        }
+
+        return Ok(None);
     }
 
-    fn tables<'a>(&'a self, catalog: &'a str, schema: &'a str) -> BoxStream<'a, Result<TableRef>> {
+    fn tables<'a>(
+        &'a self,
+        catalog: &'a str,
+        schema: &'a str,
+        query_ctx: Option<&'a QueryContext>,
+    ) -> BoxStream<'a, Result<TableRef>> {
         let sys_tables = try_stream!({
             // System tables
-            let sys_table_names = self.system_catalog.table_names(schema);
+            let sys_table_names = self.system_catalog.table_names(schema, query_ctx);
             for table_name in sys_table_names {
-                if let Some(table) = self.system_catalog.table(catalog, schema, &table_name) {
+                if let Some(table) =
+                    self.system_catalog
+                        .table(catalog, schema, &table_name, query_ctx)
+                {
                     yield table;
                 }
             }
         });
 
-        let table_id_stream = self
-            .table_metadata_manager
-            .table_name_manager()
-            .tables(catalog, schema)
-            .map_ok(|(_, v)| v.table_id());
         const BATCH_SIZE: usize = 128;
-        let user_tables = try_stream!({
+        const CONCURRENCY: usize = 8;
+
+        let (tx, rx) = tokio::sync::mpsc::channel(64);
+        let metadata_manager = self.table_metadata_manager.clone();
+        let catalog = catalog.to_string();
+        let schema = schema.to_string();
+        let semaphore = Arc::new(Semaphore::new(CONCURRENCY));
+
+        common_runtime::spawn_global(async move {
+            let table_id_stream = metadata_manager
+                .table_name_manager()
+                .tables(&catalog, &schema)
+                .map_ok(|(_, v)| v.table_id());
             // Split table ids into chunks
             let mut table_id_chunks = table_id_stream.ready_chunks(BATCH_SIZE);
 
             while let Some(table_ids) = table_id_chunks.next().await {
-                let table_ids = table_ids
+                let table_ids = match table_ids
                     .into_iter()
                     .collect::<std::result::Result<Vec<_>, _>>()
                     .map_err(BoxedError::new)
-                    .context(ListTablesSnafu { catalog, schema })?;
+                    .context(ListTablesSnafu {
+                        catalog: &catalog,
+                        schema: &schema,
+                    }) {
+                    Ok(table_ids) => table_ids,
+                    Err(e) => {
+                        let _ = tx.send(Err(e)).await;
+                        return;
+                    }
+                };
 
-                let table_info_values = self
-                    .table_metadata_manager
-                    .table_info_manager()
-                    .batch_get(&table_ids)
-                    .await
-                    .context(TableMetadataManagerSnafu)?;
+                let metadata_manager = metadata_manager.clone();
+                let tx = tx.clone();
+                let semaphore = semaphore.clone();
+                common_runtime::spawn_global(async move {
+                    // we don't explicitly close the semaphore so just ignore the potential error.
+                    let _ = semaphore.acquire().await;
+                    let table_info_values = match metadata_manager
+                        .table_info_manager()
+                        .batch_get(&table_ids)
+                        .await
+                        .context(TableMetadataManagerSnafu)
+                    {
+                        Ok(table_info_values) => table_info_values,
+                        Err(e) => {
+                            let _ = tx.send(Err(e)).await;
+                            return;
+                        }
+                    };
 
-                for table_info_value in table_info_values.into_values() {
-                    yield build_table(table_info_value)?;
-                }
+                    for table in table_info_values.into_values().map(build_table) {
+                        if tx.send(table).await.is_err() {
+                            return;
+                        }
+                    }
+                });
             }
         });
 
+        let user_tables = ReceiverStream::new(rx);
         Box::pin(sys_tables.chain(user_tables))
     }
 }
@@ -313,25 +392,34 @@ struct SystemCatalog {
     catalog_cache: Cache<String, Arc<InformationSchemaProvider>>,
     pg_catalog_cache: Cache<String, Arc<PGCatalogProvider>>,
 
-    // system_schema_provier for default catalog
+    // system_schema_provider for default catalog
     information_schema_provider: Arc<InformationSchemaProvider>,
     pg_catalog_provider: Arc<PGCatalogProvider>,
     backend: KvBackendRef,
 }
 
 impl SystemCatalog {
-    // TODO(j0hn50n133): remove the duplicated hard-coded table names logic
-    fn schema_names(&self) -> Vec<String> {
-        vec![
-            INFORMATION_SCHEMA_NAME.to_string(),
-            PG_CATALOG_NAME.to_string(),
-        ]
+    fn schema_names(&self, query_ctx: Option<&QueryContext>) -> Vec<String> {
+        let channel = query_ctx.map_or(Channel::Unknown, |ctx| ctx.channel());
+        match channel {
+            // pg_catalog only visible under postgres protocol
+            Channel::Postgres => vec![
+                INFORMATION_SCHEMA_NAME.to_string(),
+                PG_CATALOG_NAME.to_string(),
+            ],
+            _ => {
+                vec![INFORMATION_SCHEMA_NAME.to_string()]
+            }
+        }
     }
 
-    fn table_names(&self, schema: &str) -> Vec<String> {
+    fn table_names(&self, schema: &str, query_ctx: Option<&QueryContext>) -> Vec<String> {
+        let channel = query_ctx.map_or(Channel::Unknown, |ctx| ctx.channel());
         match schema {
             INFORMATION_SCHEMA_NAME => self.information_schema_provider.table_names(),
-            PG_CATALOG_NAME => self.pg_catalog_provider.table_names(),
+            PG_CATALOG_NAME if channel == Channel::Postgres => {
+                self.pg_catalog_provider.table_names()
+            }
             DEFAULT_SCHEMA_NAME => {
                 vec![NUMBERS_TABLE_NAME.to_string()]
             }
@@ -339,23 +427,35 @@ impl SystemCatalog {
         }
     }
 
-    fn schema_exists(&self, schema: &str) -> bool {
-        schema == INFORMATION_SCHEMA_NAME || schema == PG_CATALOG_NAME
+    fn schema_exists(&self, schema: &str, query_ctx: Option<&QueryContext>) -> bool {
+        let channel = query_ctx.map_or(Channel::Unknown, |ctx| ctx.channel());
+        match channel {
+            Channel::Postgres => schema == PG_CATALOG_NAME || schema == INFORMATION_SCHEMA_NAME,
+            _ => schema == INFORMATION_SCHEMA_NAME,
+        }
     }
 
-    fn table_exists(&self, schema: &str, table: &str) -> bool {
+    fn table_exists(&self, schema: &str, table: &str, query_ctx: Option<&QueryContext>) -> bool {
+        let channel = query_ctx.map_or(Channel::Unknown, |ctx| ctx.channel());
         if schema == INFORMATION_SCHEMA_NAME {
             self.information_schema_provider.table(table).is_some()
         } else if schema == DEFAULT_SCHEMA_NAME {
             table == NUMBERS_TABLE_NAME
-        } else if schema == PG_CATALOG_NAME {
+        } else if schema == PG_CATALOG_NAME && channel == Channel::Postgres {
             self.pg_catalog_provider.table(table).is_some()
         } else {
             false
         }
     }
 
-    fn table(&self, catalog: &str, schema: &str, table_name: &str) -> Option<TableRef> {
+    fn table(
+        &self,
+        catalog: &str,
+        schema: &str,
+        table_name: &str,
+        query_ctx: Option<&QueryContext>,
+    ) -> Option<TableRef> {
+        let channel = query_ctx.map_or(Channel::Unknown, |ctx| ctx.channel());
         if schema == INFORMATION_SCHEMA_NAME {
             let information_schema_provider =
                 self.catalog_cache.get_with_by_ref(catalog, move || {
@@ -366,7 +466,7 @@ impl SystemCatalog {
                     ))
                 });
             information_schema_provider.table(table_name)
-        } else if schema == PG_CATALOG_NAME {
+        } else if schema == PG_CATALOG_NAME && channel == Channel::Postgres {
             if catalog == DEFAULT_CATALOG_NAME {
                 self.pg_catalog_provider.table(table_name)
             } else {

src/catalog/src/lib.rs

@@ -20,14 +20,17 @@ use std::fmt::{Debug, Formatter};
 use std::sync::Arc;
 
 use api::v1::CreateTableExpr;
+use common_catalog::consts::{INFORMATION_SCHEMA_NAME, PG_CATALOG_NAME};
 use futures::future::BoxFuture;
 use futures_util::stream::BoxStream;
+use session::context::QueryContext;
 use table::metadata::TableId;
 use table::TableRef;
 
 use crate::error::Result;
 
 pub mod error;
+pub mod information_extension;
 pub mod kvbackend;
 pub mod memory;
 mod metrics;
@@ -44,15 +47,35 @@ pub trait CatalogManager: Send + Sync {
 
     async fn catalog_names(&self) -> Result<Vec<String>>;
 
-    async fn schema_names(&self, catalog: &str) -> Result<Vec<String>>;
+    async fn schema_names(
+        &self,
+        catalog: &str,
+        query_ctx: Option<&QueryContext>,
+    ) -> Result<Vec<String>>;
 
-    async fn table_names(&self, catalog: &str, schema: &str) -> Result<Vec<String>>;
+    async fn table_names(
+        &self,
+        catalog: &str,
+        schema: &str,
+        query_ctx: Option<&QueryContext>,
+    ) -> Result<Vec<String>>;
 
     async fn catalog_exists(&self, catalog: &str) -> Result<bool>;
 
-    async fn schema_exists(&self, catalog: &str, schema: &str) -> Result<bool>;
+    async fn schema_exists(
+        &self,
+        catalog: &str,
+        schema: &str,
+        query_ctx: Option<&QueryContext>,
+    ) -> Result<bool>;
 
-    async fn table_exists(&self, catalog: &str, schema: &str, table: &str) -> Result<bool>;
+    async fn table_exists(
+        &self,
+        catalog: &str,
+        schema: &str,
+        table: &str,
+        query_ctx: Option<&QueryContext>,
+    ) -> Result<bool>;
 
     /// Returns the table by catalog, schema and table name.
     async fn table(
@@ -60,10 +83,25 @@ pub trait CatalogManager: Send + Sync {
         catalog: &str,
         schema: &str,
         table_name: &str,
+        query_ctx: Option<&QueryContext>,
     ) -> Result<Option<TableRef>>;
 
     /// Returns all tables with a stream by catalog and schema.
-    fn tables<'a>(&'a self, catalog: &'a str, schema: &'a str) -> BoxStream<'a, Result<TableRef>>;
+    fn tables<'a>(
+        &'a self,
+        catalog: &'a str,
+        schema: &'a str,
+        query_ctx: Option<&'a QueryContext>,
+    ) -> BoxStream<'a, Result<TableRef>>;
+
+    /// Check if `schema` is a reserved schema name
+    fn is_reserved_schema_name(&self, schema: &str) -> bool {
+        // We have to check whether a schema name is reserved before create schema.
+        // We need this rather than use schema_exists directly because `pg_catalog` is
+        // only visible via postgres protocol. So if we don't check, a mysql client may
+        // create a schema named `pg_catalog` which is somehow malformed.
+        schema == INFORMATION_SCHEMA_NAME || schema == PG_CATALOG_NAME
+    }
 }
 
 pub type CatalogManagerRef = Arc<dyn CatalogManager>;

src/catalog/src/memory/manager.rs

@@ -26,6 +26,7 @@ use common_catalog::consts::{
 use common_meta::key::flow::FlowMetadataManager;
 use common_meta::kv_backend::memory::MemoryKvBackend;
 use futures_util::stream::BoxStream;
+use session::context::QueryContext;
 use snafu::OptionExt;
 use table::TableRef;
 
@@ -53,7 +54,11 @@ impl CatalogManager for MemoryCatalogManager {
         Ok(self.catalogs.read().unwrap().keys().cloned().collect())
     }
 
-    async fn schema_names(&self, catalog: &str) -> Result<Vec<String>> {
+    async fn schema_names(
+        &self,
+        catalog: &str,
+        _query_ctx: Option<&QueryContext>,
+    ) -> Result<Vec<String>> {
         Ok(self
             .catalogs
             .read()
@@ -67,7 +72,12 @@ impl CatalogManager for MemoryCatalogManager {
             .collect())
     }
 
-    async fn table_names(&self, catalog: &str, schema: &str) -> Result<Vec<String>> {
+    async fn table_names(
+        &self,
+        catalog: &str,
+        schema: &str,
+        _query_ctx: Option<&QueryContext>,
+    ) -> Result<Vec<String>> {
         Ok(self
             .catalogs
             .read()
@@ -87,11 +97,22 @@ impl CatalogManager for MemoryCatalogManager {
         self.catalog_exist_sync(catalog)
     }
 
-    async fn schema_exists(&self, catalog: &str, schema: &str) -> Result<bool> {
+    async fn schema_exists(
+        &self,
+        catalog: &str,
+        schema: &str,
+        _query_ctx: Option<&QueryContext>,
+    ) -> Result<bool> {
         self.schema_exist_sync(catalog, schema)
     }
 
-    async fn table_exists(&self, catalog: &str, schema: &str, table: &str) -> Result<bool> {
+    async fn table_exists(
+        &self,
+        catalog: &str,
+        schema: &str,
+        table: &str,
+        _query_ctx: Option<&QueryContext>,
+    ) -> Result<bool> {
         let catalogs = self.catalogs.read().unwrap();
         Ok(catalogs
             .get(catalog)
@@ -108,6 +129,7 @@ impl CatalogManager for MemoryCatalogManager {
         catalog: &str,
         schema: &str,
         table_name: &str,
+        _query_ctx: Option<&QueryContext>,
     ) -> Result<Option<TableRef>> {
         let result = try {
             self.catalogs
@@ -121,7 +143,12 @@ impl CatalogManager for MemoryCatalogManager {
         Ok(result)
     }
 
-    fn tables<'a>(&'a self, catalog: &'a str, schema: &'a str) -> BoxStream<'a, Result<TableRef>> {
+    fn tables<'a>(
+        &'a self,
+        catalog: &'a str,
+        schema: &'a str,
+        _query_ctx: Option<&QueryContext>,
+    ) -> BoxStream<'a, Result<TableRef>> {
         let catalogs = self.catalogs.read().unwrap();
 
         let Some(schemas) = catalogs.get(catalog) else {
@@ -371,11 +398,12 @@ mod tests {
                 DEFAULT_CATALOG_NAME,
                 DEFAULT_SCHEMA_NAME,
                 NUMBERS_TABLE_NAME,
+                None,
             )
             .await
             .unwrap()
             .unwrap();
-        let stream = catalog_list.tables(DEFAULT_CATALOG_NAME, DEFAULT_SCHEMA_NAME);
+        let stream = catalog_list.tables(DEFAULT_CATALOG_NAME, DEFAULT_SCHEMA_NAME, None);
         let tables = stream.try_collect::<Vec<_>>().await.unwrap();
         assert_eq!(tables.len(), 1);
         assert_eq!(
@@ -384,7 +412,12 @@ mod tests {
         );
 
         assert!(catalog_list
-            .table(DEFAULT_CATALOG_NAME, DEFAULT_SCHEMA_NAME, "not_exists")
+            .table(
+                DEFAULT_CATALOG_NAME,
+                DEFAULT_SCHEMA_NAME,
+                "not_exists",
+                None
+            )
             .await
             .unwrap()
             .is_none());
@@ -411,7 +444,7 @@ mod tests {
         };
         catalog.register_table_sync(register_table_req).unwrap();
         assert!(catalog
-            .table(DEFAULT_CATALOG_NAME, DEFAULT_SCHEMA_NAME, table_name)
+            .table(DEFAULT_CATALOG_NAME, DEFAULT_SCHEMA_NAME, table_name, None)
             .await
             .unwrap()
             .is_some());
@@ -423,7 +456,7 @@ mod tests {
         };
         catalog.deregister_table_sync(deregister_table_req).unwrap();
         assert!(catalog
-            .table(DEFAULT_CATALOG_NAME, DEFAULT_SCHEMA_NAME, table_name)
+            .table(DEFAULT_CATALOG_NAME, DEFAULT_SCHEMA_NAME, table_name, None)
             .await
             .unwrap()
             .is_none());

src/catalog/src/system_schema/information_schema.rs

@@ -18,7 +18,9 @@ pub mod flows;
 mod information_memory_table;
 pub mod key_column_usage;
 mod partitions;
+mod procedure_info;
 mod region_peers;
+mod region_statistics;
 mod runtime_metrics;
 pub mod schemata;
 mod table_constraints;
@@ -30,7 +32,11 @@ use std::collections::HashMap;
 use std::sync::{Arc, Weak};
 
 use common_catalog::consts::{self, DEFAULT_CATALOG_NAME, INFORMATION_SCHEMA_NAME};
+use common_error::ext::ErrorExt;
+use common_meta::cluster::NodeInfo;
+use common_meta::datanode::RegionStat;
 use common_meta::key::flow::FlowMetadataManager;
+use common_procedure::ProcedureInfo;
 use common_recordbatch::SendableRecordBatchStream;
 use datatypes::schema::SchemaRef;
 use lazy_static::lazy_static;
@@ -43,7 +49,7 @@ use views::InformationSchemaViews;
 
 use self::columns::InformationSchemaColumns;
 use super::{SystemSchemaProviderInner, SystemTable, SystemTableRef};
-use crate::error::Result;
+use crate::error::{Error, Result};
 use crate::system_schema::information_schema::cluster_info::InformationSchemaClusterInfo;
 use crate::system_schema::information_schema::flows::InformationSchemaFlows;
 use crate::system_schema::information_schema::information_memory_table::get_schema_columns;
@@ -188,6 +194,16 @@ impl SystemSchemaProviderInner for InformationSchemaProvider {
                 self.catalog_name.clone(),
                 self.flow_metadata_manager.clone(),
             )) as _),
+            PROCEDURE_INFO => Some(
+                Arc::new(procedure_info::InformationSchemaProcedureInfo::new(
+                    self.catalog_manager.clone(),
+                )) as _,
+            ),
+            REGION_STATISTICS => Some(Arc::new(
+                region_statistics::InformationSchemaRegionStatistics::new(
+                    self.catalog_manager.clone(),
+                ),
+            ) as _),
             _ => None,
         }
     }
@@ -235,6 +251,14 @@ impl InformationSchemaProvider {
                 CLUSTER_INFO.to_string(),
                 self.build_table(CLUSTER_INFO).unwrap(),
             );
+            tables.insert(
+                PROCEDURE_INFO.to_string(),
+                self.build_table(PROCEDURE_INFO).unwrap(),
+            );
+            tables.insert(
+                REGION_STATISTICS.to_string(),
+                self.build_table(REGION_STATISTICS).unwrap(),
+            );
         }
 
         tables.insert(TABLES.to_string(), self.build_table(TABLES).unwrap());
@@ -250,7 +274,6 @@ impl InformationSchemaProvider {
             self.build_table(TABLE_CONSTRAINTS).unwrap(),
         );
         tables.insert(FLOWS.to_string(), self.build_table(FLOWS).unwrap());
-
         // Add memory tables
         for name in MEMORY_TABLES.iter() {
             tables.insert((*name).to_string(), self.build_table(name).expect(name));
@@ -299,3 +322,39 @@ where
         InformationTable::to_stream(self, request)
     }
 }
+
+pub type InformationExtensionRef = Arc<dyn InformationExtension<Error = Error> + Send + Sync>;
+
+/// The `InformationExtension` trait provides the extension methods for the `information_schema` tables.
+#[async_trait::async_trait]
+pub trait InformationExtension {
+    type Error: ErrorExt;
+
+    /// Gets the nodes information.
+    async fn nodes(&self) -> std::result::Result<Vec<NodeInfo>, Self::Error>;
+
+    /// Gets the procedures information.
+    async fn procedures(&self) -> std::result::Result<Vec<(String, ProcedureInfo)>, Self::Error>;
+
+    /// Gets the region statistics.
+    async fn region_stats(&self) -> std::result::Result<Vec<RegionStat>, Self::Error>;
+}
+
+pub struct NoopInformationExtension;
+
+#[async_trait::async_trait]
+impl InformationExtension for NoopInformationExtension {
+    type Error = Error;
+
+    async fn nodes(&self) -> std::result::Result<Vec<NodeInfo>, Self::Error> {
+        Ok(vec![])
+    }
+
+    async fn procedures(&self) -> std::result::Result<Vec<(String, ProcedureInfo)>, Self::Error> {
+        Ok(vec![])
+    }
+
+    async fn region_stats(&self) -> std::result::Result<Vec<RegionStat>, Self::Error> {
+        Ok(vec![])
+    }
+}

src/catalog/src/system_schema/information_schema/cluster_info.rs

@@ -17,13 +17,10 @@ use std::time::Duration;
 
 use arrow_schema::SchemaRef as ArrowSchemaRef;
 use common_catalog::consts::INFORMATION_SCHEMA_CLUSTER_INFO_TABLE_ID;
-use common_config::Mode;
 use common_error::ext::BoxedError;
-use common_meta::cluster::{ClusterInfo, NodeInfo, NodeStatus};
-use common_meta::peer::Peer;
+use common_meta::cluster::NodeInfo;
 use common_recordbatch::adapter::RecordBatchStreamAdapter;
 use common_recordbatch::{RecordBatch, SendableRecordBatchStream};
-use common_telemetry::warn;
 use common_time::timestamp::Timestamp;
 use datafusion::execution::TaskContext;
 use datafusion::physical_plan::stream::RecordBatchStreamAdapter as DfRecordBatchStreamAdapter;
@@ -40,7 +37,7 @@ use snafu::ResultExt;
 use store_api::storage::{ScanRequest, TableId};
 
 use super::CLUSTER_INFO;
-use crate::error::{CreateRecordBatchSnafu, InternalSnafu, ListNodesSnafu, Result};
+use crate::error::{CreateRecordBatchSnafu, InternalSnafu, Result};
 use crate::system_schema::information_schema::{InformationTable, Predicates};
 use crate::system_schema::utils;
 use crate::CatalogManager;
@@ -70,7 +67,6 @@ const INIT_CAPACITY: usize = 42;
 pub(super) struct InformationSchemaClusterInfo {
     schema: SchemaRef,
     catalog_manager: Weak<dyn CatalogManager>,
-    start_time_ms: u64,
 }
 
 impl InformationSchemaClusterInfo {
@@ -78,7 +74,6 @@ impl InformationSchemaClusterInfo {
         Self {
             schema: Self::schema(),
             catalog_manager,
-            start_time_ms: common_time::util::current_time_millis() as u64,
         }
     }
 
@@ -100,11 +95,7 @@ impl InformationSchemaClusterInfo {
     }
 
     fn builder(&self) -> InformationSchemaClusterInfoBuilder {
-        InformationSchemaClusterInfoBuilder::new(
-            self.schema.clone(),
-            self.catalog_manager.clone(),
-            self.start_time_ms,
-        )
+        InformationSchemaClusterInfoBuilder::new(self.schema.clone(), self.catalog_manager.clone())
     }
 }
 
@@ -144,7 +135,6 @@ impl InformationTable for InformationSchemaClusterInfo {
 
 struct InformationSchemaClusterInfoBuilder {
     schema: SchemaRef,
-    start_time_ms: u64,
     catalog_manager: Weak<dyn CatalogManager>,
 
     peer_ids: Int64VectorBuilder,
@@ -158,11 +148,7 @@ struct InformationSchemaClusterInfoBuilder {
 }
 
 impl InformationSchemaClusterInfoBuilder {
-    fn new(
-        schema: SchemaRef,
-        catalog_manager: Weak<dyn CatalogManager>,
-        start_time_ms: u64,
-    ) -> Self {
+    fn new(schema: SchemaRef, catalog_manager: Weak<dyn CatalogManager>) -> Self {
         Self {
             schema,
             catalog_manager,
@@ -174,56 +160,17 @@ impl InformationSchemaClusterInfoBuilder {
             start_times: TimestampMillisecondVectorBuilder::with_capacity(INIT_CAPACITY),
             uptimes: StringVectorBuilder::with_capacity(INIT_CAPACITY),
             active_times: StringVectorBuilder::with_capacity(INIT_CAPACITY),
-            start_time_ms,
         }
     }
 
     /// Construct the `information_schema.cluster_info` virtual table
     async fn make_cluster_info(&mut self, request: Option<ScanRequest>) -> Result<RecordBatch> {
         let predicates = Predicates::from_scan_request(&request);
-        let mode = utils::running_mode(&self.catalog_manager)?.unwrap_or(Mode::Standalone);
-
-        match mode {
-            Mode::Standalone => {
-                let build_info = common_version::build_info();
-
-                self.add_node_info(
-                    &predicates,
-                    NodeInfo {
-                        // For the standalone:
-                        // - id always 0
-                        // - empty string for peer_addr
-                        peer: Peer {
-                            id: 0,
-                            addr: "".to_string(),
-                        },
-                        last_activity_ts: -1,
-                        status: NodeStatus::Standalone,
-                        version: build_info.version.to_string(),
-                        git_commit: build_info.commit_short.to_string(),
-                        // Use `self.start_time_ms` instead.
-                        // It's not precise but enough.
-                        start_time_ms: self.start_time_ms,
-                    },
-                );
-            }
-            Mode::Distributed => {
-                if let Some(meta_client) = utils::meta_client(&self.catalog_manager)? {
-                    let node_infos = meta_client
-                        .list_nodes(None)
-                        .await
-                        .map_err(BoxedError::new)
-                        .context(ListNodesSnafu)?;
-
-                    for node_info in node_infos {
-                        self.add_node_info(&predicates, node_info);
-                    }
-                } else {
-                    warn!("Could not find meta client in distributed mode.");
-                }
-            }
+        let information_extension = utils::information_extension(&self.catalog_manager)?;
+        let node_infos = information_extension.nodes().await?;
+        for node_info in node_infos {
+            self.add_node_info(&predicates, node_info);
         }
-
         self.finish()
     }
 

src/catalog/src/system_schema/information_schema/columns.rs

@@ -257,8 +257,8 @@ impl InformationSchemaColumnsBuilder {
             .context(UpgradeWeakCatalogManagerRefSnafu)?;
         let predicates = Predicates::from_scan_request(&request);
 
-        for schema_name in catalog_manager.schema_names(&catalog_name).await? {
-            let mut stream = catalog_manager.tables(&catalog_name, &schema_name);
+        for schema_name in catalog_manager.schema_names(&catalog_name, None).await? {
+            let mut stream = catalog_manager.tables(&catalog_name, &schema_name, None);
 
             while let Some(table) = stream.try_next().await? {
                 let keys = &table.table_info().meta.primary_key_indices;

src/catalog/src/system_schema/information_schema/key_column_usage.rs

@@ -212,8 +212,8 @@ impl InformationSchemaKeyColumnUsageBuilder {
             .context(UpgradeWeakCatalogManagerRefSnafu)?;
         let predicates = Predicates::from_scan_request(&request);
 
-        for schema_name in catalog_manager.schema_names(&catalog_name).await? {
-            let mut stream = catalog_manager.tables(&catalog_name, &schema_name);
+        for schema_name in catalog_manager.schema_names(&catalog_name, None).await? {
+            let mut stream = catalog_manager.tables(&catalog_name, &schema_name, None);
 
             while let Some(table) = stream.try_next().await? {
                 let mut primary_constraints = vec![];

src/catalog/src/system_schema/information_schema/partitions.rs

@@ -34,15 +34,14 @@ use datatypes::vectors::{
 };
 use futures::{StreamExt, TryStreamExt};
 use partition::manager::PartitionInfo;
-use partition::partition::PartitionDef;
 use snafu::{OptionExt, ResultExt};
-use store_api::storage::{RegionId, ScanRequest, TableId};
+use store_api::storage::{ScanRequest, TableId};
 use table::metadata::{TableInfo, TableType};
 
 use super::PARTITIONS;
 use crate::error::{
-    CreateRecordBatchSnafu, FindPartitionsSnafu, InternalSnafu, Result,
-    UpgradeWeakCatalogManagerRefSnafu,
+    CreateRecordBatchSnafu, FindPartitionsSnafu, InternalSnafu, PartitionManagerNotFoundSnafu,
+    Result, UpgradeWeakCatalogManagerRefSnafu,
 };
 use crate::kvbackend::KvBackendCatalogManager;
 use crate::system_schema::information_schema::{InformationTable, Predicates};
@@ -236,13 +235,14 @@ impl InformationSchemaPartitionsBuilder {
         let partition_manager = catalog_manager
             .as_any()
             .downcast_ref::<KvBackendCatalogManager>()
-            .map(|catalog_manager| catalog_manager.partition_manager());
+            .map(|catalog_manager| catalog_manager.partition_manager())
+            .context(PartitionManagerNotFoundSnafu)?;
 
         let predicates = Predicates::from_scan_request(&request);
 
-        for schema_name in catalog_manager.schema_names(&catalog_name).await? {
+        for schema_name in catalog_manager.schema_names(&catalog_name, None).await? {
             let table_info_stream = catalog_manager
-                .tables(&catalog_name, &schema_name)
+                .tables(&catalog_name, &schema_name, None)
                 .try_filter_map(|t| async move {
                     let table_info = t.table_info();
                     if table_info.table_type == TableType::Temporary {
@@ -262,27 +262,10 @@ impl InformationSchemaPartitionsBuilder {
                 let table_ids: Vec<TableId> =
                     table_infos.iter().map(|info| info.ident.table_id).collect();
 
-                let mut table_partitions = if let Some(partition_manager) = &partition_manager {
-                    partition_manager
-                        .batch_find_table_partitions(&table_ids)
-                        .await
-                        .context(FindPartitionsSnafu)?
-                } else {
-                    // Current node must be a standalone instance, contains only one partition by default.
-                    // TODO(dennis): change it when we support multi-regions for standalone.
-                    table_ids
-                        .into_iter()
-                        .map(|table_id| {
-                            (
-                                table_id,
-                                vec![PartitionInfo {
-                                    id: RegionId::new(table_id, 0),
-                                    partition: PartitionDef::new(vec![], vec![]),
-                                }],
-                            )
-                        })
-                        .collect()
-                };
+                let mut table_partitions = partition_manager
+                    .batch_find_table_partitions(&table_ids)
+                    .await
+                    .context(FindPartitionsSnafu)?;
 
                 for table_info in table_infos {
                     let partitions = table_partitions

src/catalog/src/system_schema/information_schema/procedure_info.rs

@@ -0,0 +1,241 @@
+// Copyright 2023 Greptime Team
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//     http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+use std::sync::{Arc, Weak};
+
+use arrow_schema::SchemaRef as ArrowSchemaRef;
+use common_catalog::consts::INFORMATION_SCHEMA_PROCEDURE_INFO_TABLE_ID;
+use common_error::ext::BoxedError;
+use common_procedure::ProcedureInfo;
+use common_recordbatch::adapter::RecordBatchStreamAdapter;
+use common_recordbatch::{RecordBatch, SendableRecordBatchStream};
+use common_time::timestamp::Timestamp;
+use datafusion::execution::TaskContext;
+use datafusion::physical_plan::stream::RecordBatchStreamAdapter as DfRecordBatchStreamAdapter;
+use datafusion::physical_plan::streaming::PartitionStream as DfPartitionStream;
+use datafusion::physical_plan::SendableRecordBatchStream as DfSendableRecordBatchStream;
+use datatypes::prelude::{ConcreteDataType, ScalarVectorBuilder, VectorRef};
+use datatypes::schema::{ColumnSchema, Schema, SchemaRef};
+use datatypes::timestamp::TimestampMillisecond;
+use datatypes::value::Value;
+use datatypes::vectors::{StringVectorBuilder, TimestampMillisecondVectorBuilder};
+use snafu::ResultExt;
+use store_api::storage::{ScanRequest, TableId};
+
+use super::PROCEDURE_INFO;
+use crate::error::{CreateRecordBatchSnafu, InternalSnafu, Result};
+use crate::system_schema::information_schema::{InformationTable, Predicates};
+use crate::system_schema::utils;
+use crate::CatalogManager;
+
+const PROCEDURE_ID: &str = "procedure_id";
+const PROCEDURE_TYPE: &str = "procedure_type";
+const START_TIME: &str = "start_time";
+const END_TIME: &str = "end_time";
+const STATUS: &str = "status";
+const LOCK_KEYS: &str = "lock_keys";
+
+const INIT_CAPACITY: usize = 42;
+
+/// The `PROCEDURE_INFO` table provides information about the current procedure information of the cluster.
+///
+/// - `procedure_id`: the unique identifier of the procedure.
+/// - `procedure_name`: the name of the procedure.
+/// - `start_time`: the starting execution time of the procedure.
+/// - `end_time`: the ending execution time of the procedure.
+/// - `status`: the status of the procedure.
+/// - `lock_keys`: the lock keys of the procedure.
+///
+pub(super) struct InformationSchemaProcedureInfo {
+    schema: SchemaRef,
+    catalog_manager: Weak<dyn CatalogManager>,
+}
+
+impl InformationSchemaProcedureInfo {
+    pub(super) fn new(catalog_manager: Weak<dyn CatalogManager>) -> Self {
+        Self {
+            schema: Self::schema(),
+            catalog_manager,
+        }
+    }
+
+    pub(crate) fn schema() -> SchemaRef {
+        Arc::new(Schema::new(vec![
+            ColumnSchema::new(PROCEDURE_ID, ConcreteDataType::string_datatype(), false),
+            ColumnSchema::new(PROCEDURE_TYPE, ConcreteDataType::string_datatype(), false),
+            ColumnSchema::new(
+                START_TIME,
+                ConcreteDataType::timestamp_millisecond_datatype(),
+                true,
+            ),
+            ColumnSchema::new(
+                END_TIME,
+                ConcreteDataType::timestamp_millisecond_datatype(),
+                true,
+            ),
+            ColumnSchema::new(STATUS, ConcreteDataType::string_datatype(), false),
+            ColumnSchema::new(LOCK_KEYS, ConcreteDataType::string_datatype(), true),
+        ]))
+    }
+
+    fn builder(&self) -> InformationSchemaProcedureInfoBuilder {
+        InformationSchemaProcedureInfoBuilder::new(
+            self.schema.clone(),
+            self.catalog_manager.clone(),
+        )
+    }
+}
+
+impl InformationTable for InformationSchemaProcedureInfo {
+    fn table_id(&self) -> TableId {
+        INFORMATION_SCHEMA_PROCEDURE_INFO_TABLE_ID
+    }
+
+    fn table_name(&self) -> &'static str {
+        PROCEDURE_INFO
+    }
+
+    fn schema(&self) -> SchemaRef {
+        self.schema.clone()
+    }
+
+    fn to_stream(&self, request: ScanRequest) -> Result<SendableRecordBatchStream> {
+        let schema = self.schema.arrow_schema().clone();
+        let mut builder = self.builder();
+        let stream = Box::pin(DfRecordBatchStreamAdapter::new(
+            schema,
+            futures::stream::once(async move {
+                builder
+                    .make_procedure_info(Some(request))
+                    .await
+                    .map(|x| x.into_df_record_batch())
+                    .map_err(Into::into)
+            }),
+        ));
+        Ok(Box::pin(
+            RecordBatchStreamAdapter::try_new(stream)
+                .map_err(BoxedError::new)
+                .context(InternalSnafu)?,
+        ))
+    }
+}
+
+struct InformationSchemaProcedureInfoBuilder {
+    schema: SchemaRef,
+    catalog_manager: Weak<dyn CatalogManager>,
+
+    procedure_ids: StringVectorBuilder,
+    procedure_types: StringVectorBuilder,
+    start_times: TimestampMillisecondVectorBuilder,
+    end_times: TimestampMillisecondVectorBuilder,
+    statuses: StringVectorBuilder,
+    lock_keys: StringVectorBuilder,
+}
+
+impl InformationSchemaProcedureInfoBuilder {
+    fn new(schema: SchemaRef, catalog_manager: Weak<dyn CatalogManager>) -> Self {
+        Self {
+            schema,
+            catalog_manager,
+            procedure_ids: StringVectorBuilder::with_capacity(INIT_CAPACITY),
+            procedure_types: StringVectorBuilder::with_capacity(INIT_CAPACITY),
+            start_times: TimestampMillisecondVectorBuilder::with_capacity(INIT_CAPACITY),
+            end_times: TimestampMillisecondVectorBuilder::with_capacity(INIT_CAPACITY),
+            statuses: StringVectorBuilder::with_capacity(INIT_CAPACITY),
+            lock_keys: StringVectorBuilder::with_capacity(INIT_CAPACITY),
+        }
+    }
+
+    /// Construct the `information_schema.procedure_info` virtual table
+    async fn make_procedure_info(&mut self, request: Option<ScanRequest>) -> Result<RecordBatch> {
+        let predicates = Predicates::from_scan_request(&request);
+        let information_extension = utils::information_extension(&self.catalog_manager)?;
+        let procedures = information_extension.procedures().await?;
+        for (status, procedure_info) in procedures {
+            self.add_procedure(&predicates, status, procedure_info);
+        }
+        self.finish()
+    }
+
+    fn add_procedure(
+        &mut self,
+        predicates: &Predicates,
+        status: String,
+        procedure_info: ProcedureInfo,
+    ) {
+        let ProcedureInfo {
+            id,
+            type_name,
+            start_time_ms,
+            end_time_ms,
+            lock_keys,
+            ..
+        } = procedure_info;
+        let pid = id.to_string();
+        let start_time = TimestampMillisecond(Timestamp::new_millisecond(start_time_ms));
+        let end_time = TimestampMillisecond(Timestamp::new_millisecond(end_time_ms));
+        let lock_keys = lock_keys.join(",");
+
+        let row = [
+            (PROCEDURE_ID, &Value::from(pid.clone())),
+            (PROCEDURE_TYPE, &Value::from(type_name.clone())),
+            (START_TIME, &Value::from(start_time)),
+            (END_TIME, &Value::from(end_time)),
+            (STATUS, &Value::from(status.clone())),
+            (LOCK_KEYS, &Value::from(lock_keys.clone())),
+        ];
+        if !predicates.eval(&row) {
+            return;
+        }
+        self.procedure_ids.push(Some(&pid));
+        self.procedure_types.push(Some(&type_name));
+        self.start_times.push(Some(start_time));
+        self.end_times.push(Some(end_time));
+        self.statuses.push(Some(&status));
+        self.lock_keys.push(Some(&lock_keys));
+    }
+
+    fn finish(&mut self) -> Result<RecordBatch> {
+        let columns: Vec<VectorRef> = vec![
+            Arc::new(self.procedure_ids.finish()),
+            Arc::new(self.procedure_types.finish()),
+            Arc::new(self.start_times.finish()),
+            Arc::new(self.end_times.finish()),
+            Arc::new(self.statuses.finish()),
+            Arc::new(self.lock_keys.finish()),
+        ];
+        RecordBatch::new(self.schema.clone(), columns).context(CreateRecordBatchSnafu)
+    }
+}
+
+impl DfPartitionStream for InformationSchemaProcedureInfo {
+    fn schema(&self) -> &ArrowSchemaRef {
+        self.schema.arrow_schema()
+    }
+
+    fn execute(&self, _: Arc<TaskContext>) -> DfSendableRecordBatchStream {
+        let schema = self.schema.arrow_schema().clone();
+        let mut builder = self.builder();
+        Box::pin(DfRecordBatchStreamAdapter::new(
+            schema,
+            futures::stream::once(async move {
+                builder
+                    .make_procedure_info(None)
+                    .await
+                    .map(|x| x.into_df_record_batch())
+                    .map_err(Into::into)
+            }),
+        ))
+    }
+}

src/catalog/src/system_schema/information_schema/region_peers.rs

@@ -176,9 +176,9 @@ impl InformationSchemaRegionPeersBuilder {
 
         let predicates = Predicates::from_scan_request(&request);
 
-        for schema_name in catalog_manager.schema_names(&catalog_name).await? {
+        for schema_name in catalog_manager.schema_names(&catalog_name, None).await? {
             let table_id_stream = catalog_manager
-                .tables(&catalog_name, &schema_name)
+                .tables(&catalog_name, &schema_name, None)
                 .try_filter_map(|t| async move {
                     let table_info = t.table_info();
                     if table_info.table_type == TableType::Temporary {
@@ -224,8 +224,8 @@ impl InformationSchemaRegionPeersBuilder {
             let region_id = RegionId::new(table_id, route.region.id.region_number()).as_u64();
             let peer_id = route.leader_peer.clone().map(|p| p.id);
             let peer_addr = route.leader_peer.clone().map(|p| p.addr);
-            let status = if let Some(status) = route.leader_status {
-                Some(status.as_ref().to_string())
+            let state = if let Some(state) = route.leader_state {
+                Some(state.as_ref().to_string())
             } else {
                 // Alive by default
                 Some("ALIVE".to_string())
@@ -242,7 +242,7 @@ impl InformationSchemaRegionPeersBuilder {
             self.peer_ids.push(peer_id);
             self.peer_addrs.push(peer_addr.as_deref());
             self.is_leaders.push(Some("Yes"));
-            self.statuses.push(status.as_deref());
+            self.statuses.push(state.as_deref());
             self.down_seconds
                 .push(route.leader_down_millis().map(|m| m / 1000));
         }

src/catalog/src/system_schema/information_schema/region_statistics.rs

@@ -0,0 +1,261 @@
+// Copyright 2023 Greptime Team
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//     http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+use std::sync::{Arc, Weak};
+
+use arrow_schema::SchemaRef as ArrowSchemaRef;
+use common_catalog::consts::INFORMATION_SCHEMA_REGION_STATISTICS_TABLE_ID;
+use common_error::ext::BoxedError;
+use common_meta::datanode::RegionStat;
+use common_recordbatch::adapter::RecordBatchStreamAdapter;
+use common_recordbatch::{DfSendableRecordBatchStream, RecordBatch, SendableRecordBatchStream};
+use datafusion::execution::TaskContext;
+use datafusion::physical_plan::stream::RecordBatchStreamAdapter as DfRecordBatchStreamAdapter;
+use datafusion::physical_plan::streaming::PartitionStream as DfPartitionStream;
+use datatypes::prelude::{ConcreteDataType, ScalarVectorBuilder, VectorRef};
+use datatypes::schema::{ColumnSchema, Schema, SchemaRef};
+use datatypes::value::Value;
+use datatypes::vectors::{StringVectorBuilder, UInt32VectorBuilder, UInt64VectorBuilder};
+use snafu::ResultExt;
+use store_api::storage::{ScanRequest, TableId};
+
+use super::{InformationTable, REGION_STATISTICS};
+use crate::error::{CreateRecordBatchSnafu, InternalSnafu, Result};
+use crate::information_schema::Predicates;
+use crate::system_schema::utils;
+use crate::CatalogManager;
+
+const REGION_ID: &str = "region_id";
+const TABLE_ID: &str = "table_id";
+const REGION_NUMBER: &str = "region_number";
+const REGION_ROWS: &str = "region_rows";
+const DISK_SIZE: &str = "disk_size";
+const MEMTABLE_SIZE: &str = "memtable_size";
+const MANIFEST_SIZE: &str = "manifest_size";
+const SST_SIZE: &str = "sst_size";
+const INDEX_SIZE: &str = "index_size";
+const ENGINE: &str = "engine";
+const REGION_ROLE: &str = "region_role";
+
+const INIT_CAPACITY: usize = 42;
+
+/// The `REGION_STATISTICS` table provides information about the region statistics. Including fields:
+///
+/// - `region_id`: The region id.
+/// - `table_id`: The table id.
+/// - `region_number`: The region number.
+/// - `region_rows`: The number of rows in region.
+/// - `memtable_size`: The memtable size in bytes.
+/// - `disk_size`: The approximate disk size in bytes.
+/// - `manifest_size`: The manifest size in bytes.
+/// - `sst_size`: The sst data files size in bytes.
+/// - `index_size`: The sst index files size in bytes.
+/// - `engine`: The engine type.
+/// - `region_role`: The region role.
+///
+pub(super) struct InformationSchemaRegionStatistics {
+    schema: SchemaRef,
+    catalog_manager: Weak<dyn CatalogManager>,
+}
+
+impl InformationSchemaRegionStatistics {
+    pub(super) fn new(catalog_manager: Weak<dyn CatalogManager>) -> Self {
+        Self {
+            schema: Self::schema(),
+            catalog_manager,
+        }
+    }
+
+    pub(crate) fn schema() -> SchemaRef {
+        Arc::new(Schema::new(vec![
+            ColumnSchema::new(REGION_ID, ConcreteDataType::uint64_datatype(), false),
+            ColumnSchema::new(TABLE_ID, ConcreteDataType::uint32_datatype(), false),
+            ColumnSchema::new(REGION_NUMBER, ConcreteDataType::uint32_datatype(), false),
+            ColumnSchema::new(REGION_ROWS, ConcreteDataType::uint64_datatype(), true),
+            ColumnSchema::new(DISK_SIZE, ConcreteDataType::uint64_datatype(), true),
+            ColumnSchema::new(MEMTABLE_SIZE, ConcreteDataType::uint64_datatype(), true),
+            ColumnSchema::new(MANIFEST_SIZE, ConcreteDataType::uint64_datatype(), true),
+            ColumnSchema::new(SST_SIZE, ConcreteDataType::uint64_datatype(), true),
+            ColumnSchema::new(INDEX_SIZE, ConcreteDataType::uint64_datatype(), true),
+            ColumnSchema::new(ENGINE, ConcreteDataType::string_datatype(), true),
+            ColumnSchema::new(REGION_ROLE, ConcreteDataType::string_datatype(), true),
+        ]))
+    }
+
+    fn builder(&self) -> InformationSchemaRegionStatisticsBuilder {
+        InformationSchemaRegionStatisticsBuilder::new(
+            self.schema.clone(),
+            self.catalog_manager.clone(),
+        )
+    }
+}
+
+impl InformationTable for InformationSchemaRegionStatistics {
+    fn table_id(&self) -> TableId {
+        INFORMATION_SCHEMA_REGION_STATISTICS_TABLE_ID
+    }
+
+    fn table_name(&self) -> &'static str {
+        REGION_STATISTICS
+    }
+
+    fn schema(&self) -> SchemaRef {
+        self.schema.clone()
+    }
+
+    fn to_stream(&self, request: ScanRequest) -> Result<SendableRecordBatchStream> {
+        let schema = self.schema.arrow_schema().clone();
+        let mut builder = self.builder();
+
+        let stream = Box::pin(DfRecordBatchStreamAdapter::new(
+            schema,
+            futures::stream::once(async move {
+                builder
+                    .make_region_statistics(Some(request))
+                    .await
+                    .map(|x| x.into_df_record_batch())
+                    .map_err(Into::into)
+            }),
+        ));
+
+        Ok(Box::pin(
+            RecordBatchStreamAdapter::try_new(stream)
+                .map_err(BoxedError::new)
+                .context(InternalSnafu)?,
+        ))
+    }
+}
+
+struct InformationSchemaRegionStatisticsBuilder {
+    schema: SchemaRef,
+    catalog_manager: Weak<dyn CatalogManager>,
+
+    region_ids: UInt64VectorBuilder,
+    table_ids: UInt32VectorBuilder,
+    region_numbers: UInt32VectorBuilder,
+    region_rows: UInt64VectorBuilder,
+    disk_sizes: UInt64VectorBuilder,
+    memtable_sizes: UInt64VectorBuilder,
+    manifest_sizes: UInt64VectorBuilder,
+    sst_sizes: UInt64VectorBuilder,
+    index_sizes: UInt64VectorBuilder,
+    engines: StringVectorBuilder,
+    region_roles: StringVectorBuilder,
+}
+
+impl InformationSchemaRegionStatisticsBuilder {
+    fn new(schema: SchemaRef, catalog_manager: Weak<dyn CatalogManager>) -> Self {
+        Self {
+            schema,
+            catalog_manager,
+            region_ids: UInt64VectorBuilder::with_capacity(INIT_CAPACITY),
+            table_ids: UInt32VectorBuilder::with_capacity(INIT_CAPACITY),
+            region_numbers: UInt32VectorBuilder::with_capacity(INIT_CAPACITY),
+            region_rows: UInt64VectorBuilder::with_capacity(INIT_CAPACITY),
+            disk_sizes: UInt64VectorBuilder::with_capacity(INIT_CAPACITY),
+            memtable_sizes: UInt64VectorBuilder::with_capacity(INIT_CAPACITY),
+            manifest_sizes: UInt64VectorBuilder::with_capacity(INIT_CAPACITY),
+            sst_sizes: UInt64VectorBuilder::with_capacity(INIT_CAPACITY),
+            index_sizes: UInt64VectorBuilder::with_capacity(INIT_CAPACITY),
+            engines: StringVectorBuilder::with_capacity(INIT_CAPACITY),
+            region_roles: StringVectorBuilder::with_capacity(INIT_CAPACITY),
+        }
+    }
+
+    /// Construct a new `InformationSchemaRegionStatistics` from the collected data.
+    async fn make_region_statistics(
+        &mut self,
+        request: Option<ScanRequest>,
+    ) -> Result<RecordBatch> {
+        let predicates = Predicates::from_scan_request(&request);
+        let information_extension = utils::information_extension(&self.catalog_manager)?;
+        let region_stats = information_extension.region_stats().await?;
+        for region_stat in region_stats {
+            self.add_region_statistic(&predicates, region_stat);
+        }
+        self.finish()
+    }
+
+    fn add_region_statistic(&mut self, predicate: &Predicates, region_stat: RegionStat) {
+        let row = [
+            (REGION_ID, &Value::from(region_stat.id.as_u64())),
+            (TABLE_ID, &Value::from(region_stat.id.table_id())),
+            (REGION_NUMBER, &Value::from(region_stat.id.region_number())),
+            (REGION_ROWS, &Value::from(region_stat.num_rows)),
+            (DISK_SIZE, &Value::from(region_stat.approximate_bytes)),
+            (MEMTABLE_SIZE, &Value::from(region_stat.memtable_size)),
+            (MANIFEST_SIZE, &Value::from(region_stat.manifest_size)),
+            (SST_SIZE, &Value::from(region_stat.sst_size)),
+            (INDEX_SIZE, &Value::from(region_stat.index_size)),
+            (ENGINE, &Value::from(region_stat.engine.as_str())),
+            (REGION_ROLE, &Value::from(region_stat.role.to_string())),
+        ];
+
+        if !predicate.eval(&row) {
+            return;
+        }
+
+        self.region_ids.push(Some(region_stat.id.as_u64()));
+        self.table_ids.push(Some(region_stat.id.table_id()));
+        self.region_numbers
+            .push(Some(region_stat.id.region_number()));
+        self.region_rows.push(Some(region_stat.num_rows));
+        self.disk_sizes.push(Some(region_stat.approximate_bytes));
+        self.memtable_sizes.push(Some(region_stat.memtable_size));
+        self.manifest_sizes.push(Some(region_stat.manifest_size));
+        self.sst_sizes.push(Some(region_stat.sst_size));
+        self.index_sizes.push(Some(region_stat.index_size));
+        self.engines.push(Some(&region_stat.engine));
+        self.region_roles.push(Some(&region_stat.role.to_string()));
+    }
+
+    fn finish(&mut self) -> Result<RecordBatch> {
+        let columns: Vec<VectorRef> = vec![
+            Arc::new(self.region_ids.finish()),
+            Arc::new(self.table_ids.finish()),
+            Arc::new(self.region_numbers.finish()),
+            Arc::new(self.region_rows.finish()),
+            Arc::new(self.disk_sizes.finish()),
+            Arc::new(self.memtable_sizes.finish()),
+            Arc::new(self.manifest_sizes.finish()),
+            Arc::new(self.sst_sizes.finish()),
+            Arc::new(self.index_sizes.finish()),
+            Arc::new(self.engines.finish()),
+            Arc::new(self.region_roles.finish()),
+        ];
+
+        RecordBatch::new(self.schema.clone(), columns).context(CreateRecordBatchSnafu)
+    }
+}
+
+impl DfPartitionStream for InformationSchemaRegionStatistics {
+    fn schema(&self) -> &ArrowSchemaRef {
+        self.schema.arrow_schema()
+    }
+
+    fn execute(&self, _: Arc<TaskContext>) -> DfSendableRecordBatchStream {
+        let schema = self.schema.arrow_schema().clone();
+        let mut builder = self.builder();
+        Box::pin(DfRecordBatchStreamAdapter::new(
+            schema,
+            futures::stream::once(async move {
+                builder
+                    .make_region_statistics(None)
+                    .await
+                    .map(|x| x.into_df_record_batch())
+                    .map_err(Into::into)
+            }),
+        ))
+    }
+}

src/catalog/src/system_schema/information_schema/schemata.rs

@@ -171,7 +171,7 @@ impl InformationSchemaSchemataBuilder {
         let table_metadata_manager = utils::table_meta_manager(&self.catalog_manager)?;
         let predicates = Predicates::from_scan_request(&request);
 
-        for schema_name in catalog_manager.schema_names(&catalog_name).await? {
+        for schema_name in catalog_manager.schema_names(&catalog_name, None).await? {
             let opts = if let Some(table_metadata_manager) = &table_metadata_manager {
                 table_metadata_manager
                     .schema_manager()
@@ -180,7 +180,7 @@ impl InformationSchemaSchemataBuilder {
                     .context(TableMetadataManagerSnafu)?
                     // information_schema is not available from this
                     // table_metadata_manager and we return None
-                    .map(|schema_opts| format!("{schema_opts}"))
+                    .map(|schema_opts| format!("{}", schema_opts.into_inner()))
             } else {
                 None
             };

src/catalog/src/system_schema/information_schema/table_constraints.rs

@@ -176,8 +176,8 @@ impl InformationSchemaTableConstraintsBuilder {
             .context(UpgradeWeakCatalogManagerRefSnafu)?;
         let predicates = Predicates::from_scan_request(&request);
 
-        for schema_name in catalog_manager.schema_names(&catalog_name).await? {
-            let mut stream = catalog_manager.tables(&catalog_name, &schema_name);
+        for schema_name in catalog_manager.schema_names(&catalog_name, None).await? {
+            let mut stream = catalog_manager.tables(&catalog_name, &schema_name, None);
 
             while let Some(table) = stream.try_next().await? {
                 let keys = &table.table_info().meta.primary_key_indices;

src/catalog/src/system_schema/information_schema/table_names.rs

@@ -12,7 +12,7 @@
 // See the License for the specific language governing permissions and
 // limitations under the License.
 
-/// All table names in `information_schema`.
+//! All table names in `information_schema`.
 
 pub const TABLES: &str = "tables";
 pub const COLUMNS: &str = "columns";
@@ -45,3 +45,5 @@ pub const TABLE_CONSTRAINTS: &str = "table_constraints";
 pub const CLUSTER_INFO: &str = "cluster_info";
 pub const VIEWS: &str = "views";
 pub const FLOWS: &str = "flows";
+pub const PROCEDURE_INFO: &str = "procedure_info";
+pub const REGION_STATISTICS: &str = "region_statistics";

src/catalog/src/system_schema/information_schema/tables.rs

@@ -12,13 +12,16 @@
 // See the License for the specific language governing permissions and
 // limitations under the License.
 
+use std::collections::HashSet;
 use std::sync::{Arc, Weak};
 
 use arrow_schema::SchemaRef as ArrowSchemaRef;
-use common_catalog::consts::INFORMATION_SCHEMA_TABLES_TABLE_ID;
+use common_catalog::consts::{INFORMATION_SCHEMA_TABLES_TABLE_ID, MITO_ENGINE};
 use common_error::ext::BoxedError;
+use common_meta::datanode::RegionStat;
 use common_recordbatch::adapter::RecordBatchStreamAdapter;
 use common_recordbatch::{RecordBatch, SendableRecordBatchStream};
+use common_telemetry::error;
 use datafusion::execution::TaskContext;
 use datafusion::physical_plan::stream::RecordBatchStreamAdapter as DfRecordBatchStreamAdapter;
 use datafusion::physical_plan::streaming::PartitionStream as DfPartitionStream;
@@ -31,7 +34,7 @@ use datatypes::vectors::{
 };
 use futures::TryStreamExt;
 use snafu::{OptionExt, ResultExt};
-use store_api::storage::{ScanRequest, TableId};
+use store_api::storage::{RegionId, ScanRequest, TableId};
 use table::metadata::{TableInfo, TableType};
 
 use super::TABLES;
@@ -39,6 +42,7 @@ use crate::error::{
     CreateRecordBatchSnafu, InternalSnafu, Result, UpgradeWeakCatalogManagerRefSnafu,
 };
 use crate::system_schema::information_schema::{InformationTable, Predicates};
+use crate::system_schema::utils;
 use crate::CatalogManager;
 
 pub const TABLE_CATALOG: &str = "table_catalog";
@@ -234,17 +238,51 @@ impl InformationSchemaTablesBuilder {
             .context(UpgradeWeakCatalogManagerRefSnafu)?;
         let predicates = Predicates::from_scan_request(&request);
 
-        for schema_name in catalog_manager.schema_names(&catalog_name).await? {
-            let mut stream = catalog_manager.tables(&catalog_name, &schema_name);
+        let information_extension = utils::information_extension(&self.catalog_manager)?;
+
+        // TODO(dennis): `region_stats` API is not stable in distributed cluster because of network issue etc.
+        // But we don't want the statements such as `show tables` fail,
+        // so using `unwrap_or_else` here instead of `?` operator.
+        let region_stats = information_extension
+            .region_stats()
+            .await
+            .map_err(|e| {
+                error!(e; "Failed to call region_stats");
+                e
+            })
+            .unwrap_or_else(|_| vec![]);
+
+        for schema_name in catalog_manager.schema_names(&catalog_name, None).await? {
+            let mut stream = catalog_manager.tables(&catalog_name, &schema_name, None);
 
             while let Some(table) = stream.try_next().await? {
                 let table_info = table.table_info();
+
+                // TODO(dennis): make it working for metric engine
+                let table_region_stats =
+                    if table_info.meta.engine == MITO_ENGINE || table_info.is_physical_table() {
+                        let region_ids = table_info
+                            .meta
+                            .region_numbers
+                            .iter()
+                            .map(|n| RegionId::new(table_info.ident.table_id, *n))
+                            .collect::<HashSet<_>>();
+
+                        region_stats
+                            .iter()
+                            .filter(|stat| region_ids.contains(&stat.id))
+                            .collect::<Vec<_>>()
+                    } else {
+                        vec![]
+                    };
+
                 self.add_table(
                     &predicates,
                     &catalog_name,
                     &schema_name,
                     table_info,
                     table.table_type(),
+                    &table_region_stats,
                 );
             }
         }
@@ -260,6 +298,7 @@ impl InformationSchemaTablesBuilder {
         schema_name: &str,
         table_info: Arc<TableInfo>,
         table_type: TableType,
+        region_stats: &[&RegionStat],
     ) {
         let table_name = table_info.name.as_ref();
         let table_id = table_info.table_id();
@@ -273,7 +312,9 @@ impl InformationSchemaTablesBuilder {
 
         let row = [
             (TABLE_CATALOG, &Value::from(catalog_name)),
+            (TABLE_ID, &Value::from(table_id)),
             (TABLE_SCHEMA, &Value::from(schema_name)),
+            (ENGINE, &Value::from(engine)),
             (TABLE_NAME, &Value::from(table_name)),
             (TABLE_TYPE, &Value::from(table_type_text)),
         ];
@@ -287,21 +328,39 @@ impl InformationSchemaTablesBuilder {
         self.table_names.push(Some(table_name));
         self.table_types.push(Some(table_type_text));
         self.table_ids.push(Some(table_id));
+
+        let data_length = region_stats.iter().map(|stat| stat.sst_size).sum();
+        let table_rows = region_stats.iter().map(|stat| stat.num_rows).sum();
+        let index_length = region_stats.iter().map(|stat| stat.index_size).sum();
+
+        // It's not precise, but it is acceptable for long-term data storage.
+        let avg_row_length = if table_rows > 0 {
+            let total_data_length = data_length
+                + region_stats
+                    .iter()
+                    .map(|stat| stat.memtable_size)
+                    .sum::<u64>();
+
+            total_data_length / table_rows
+        } else {
+            0
+        };
+
+        self.data_length.push(Some(data_length));
+        self.index_length.push(Some(index_length));
+        self.table_rows.push(Some(table_rows));
+        self.avg_row_length.push(Some(avg_row_length));
+
         // TODO(sunng87): use real data for these fields
-        self.data_length.push(Some(0));
         self.max_data_length.push(Some(0));
-        self.index_length.push(Some(0));
-        self.avg_row_length.push(Some(0));
-        self.max_index_length.push(Some(0));
         self.checksum.push(Some(0));
-        self.table_rows.push(Some(0));
+        self.max_index_length.push(Some(0));
         self.data_free.push(Some(0));
         self.auto_increment.push(Some(0));
         self.row_format.push(Some("Fixed"));
         self.table_collation.push(Some("utf8_bin"));
         self.update_time.push(None);
         self.check_time.push(None);
-
         // use mariadb default table version number here
         self.version.push(Some(11));
         self.table_comment.push(table_info.desc.as_deref());

src/catalog/src/system_schema/information_schema/views.rs

@@ -192,8 +192,8 @@ impl InformationSchemaViewsBuilder {
             .context(CastManagerSnafu)?
             .view_info_cache()?;
 
-        for schema_name in catalog_manager.schema_names(&catalog_name).await? {
-            let mut stream = catalog_manager.tables(&catalog_name, &schema_name);
+        for schema_name in catalog_manager.schema_names(&catalog_name, None).await? {
+            let mut stream = catalog_manager.tables(&catalog_name, &schema_name, None);
 
             while let Some(table) = stream.try_next().await? {
                 let table_info = table.table_info();

src/catalog/src/system_schema/memory_table.rs

@@ -74,7 +74,7 @@ impl MemoryTableBuilder {
     /// Construct the `information_schema.{table_name}` virtual table
     pub async fn memory_records(&mut self) -> Result<RecordBatch> {
         if self.columns.is_empty() {
-            RecordBatch::new_empty(self.schema.clone()).context(CreateRecordBatchSnafu)
+            Ok(RecordBatch::new_empty(self.schema.clone()))
         } else {
             RecordBatch::new(self.schema.clone(), std::mem::take(&mut self.columns))
                 .context(CreateRecordBatchSnafu)

src/catalog/src/system_schema/pg_catalog.rs

@@ -18,15 +18,16 @@ mod pg_namespace;
 mod table_names;
 
 use std::collections::HashMap;
-use std::sync::{Arc, Weak};
+use std::sync::{Arc, LazyLock, Weak};
 
-use common_catalog::consts::{self, PG_CATALOG_NAME};
+use common_catalog::consts::{self, DEFAULT_CATALOG_NAME, DEFAULT_SCHEMA_NAME, PG_CATALOG_NAME};
 use datatypes::schema::ColumnSchema;
 use lazy_static::lazy_static;
 use paste::paste;
 use pg_catalog_memory_table::get_schema_columns;
 use pg_class::PGClass;
 use pg_namespace::PGNamespace;
+use session::context::{Channel, QueryContext};
 use table::TableRef;
 pub use table_names::*;
 
@@ -142,3 +143,12 @@ impl SystemSchemaProviderInner for PGCatalogProvider {
         &self.catalog_name
     }
 }
+
+/// Provide query context to call the [`CatalogManager`]'s method.
+static PG_QUERY_CTX: LazyLock<QueryContext> = LazyLock::new(|| {
+    QueryContext::with_channel(DEFAULT_CATALOG_NAME, DEFAULT_SCHEMA_NAME, Channel::Postgres)
+});
+
+fn query_ctx() -> Option<&'static QueryContext> {
+    Some(&PG_QUERY_CTX)
+}

src/catalog/src/system_schema/pg_catalog/pg_class.rs

@@ -32,7 +32,7 @@ use store_api::storage::ScanRequest;
 use table::metadata::TableType;
 
 use super::pg_namespace::oid_map::PGNamespaceOidMapRef;
-use super::{OID_COLUMN_NAME, PG_CLASS};
+use super::{query_ctx, OID_COLUMN_NAME, PG_CLASS};
 use crate::error::{
     CreateRecordBatchSnafu, InternalSnafu, Result, UpgradeWeakCatalogManagerRefSnafu,
 };
@@ -202,8 +202,11 @@ impl PGClassBuilder {
             .upgrade()
             .context(UpgradeWeakCatalogManagerRefSnafu)?;
         let predicates = Predicates::from_scan_request(&request);
-        for schema_name in catalog_manager.schema_names(&catalog_name).await? {
-            let mut stream = catalog_manager.tables(&catalog_name, &schema_name);
+        for schema_name in catalog_manager
+            .schema_names(&catalog_name, query_ctx())
+            .await?
+        {
+            let mut stream = catalog_manager.tables(&catalog_name, &schema_name, query_ctx());
             while let Some(table) = stream.try_next().await? {
                 let table_info = table.table_info();
                 self.add_class(

src/catalog/src/system_schema/pg_catalog/pg_namespace.rs

@@ -12,6 +12,9 @@
 // See the License for the specific language governing permissions and
 // limitations under the License.
 
+//! The `pg_catalog.pg_namespace` table implementation.
+//! namespace is a schema in greptime
+
 pub(super) mod oid_map;
 
 use std::sync::{Arc, Weak};
@@ -31,7 +34,7 @@ use datatypes::vectors::{StringVectorBuilder, UInt32VectorBuilder, VectorRef};
 use snafu::{OptionExt, ResultExt};
 use store_api::storage::ScanRequest;
 
-use super::{PGNamespaceOidMapRef, OID_COLUMN_NAME, PG_NAMESPACE};
+use super::{query_ctx, PGNamespaceOidMapRef, OID_COLUMN_NAME, PG_NAMESPACE};
 use crate::error::{
     CreateRecordBatchSnafu, InternalSnafu, Result, UpgradeWeakCatalogManagerRefSnafu,
 };
@@ -40,9 +43,6 @@ use crate::system_schema::utils::tables::{string_column, u32_column};
 use crate::system_schema::SystemTable;
 use crate::CatalogManager;
 
-/// The `pg_catalog.pg_namespace` table implementation.
-/// namespace is a schema in greptime
-
 const NSPNAME: &str = "nspname";
 const INIT_CAPACITY: usize = 42;
 
@@ -180,7 +180,10 @@ impl PGNamespaceBuilder {
             .upgrade()
             .context(UpgradeWeakCatalogManagerRefSnafu)?;
         let predicates = Predicates::from_scan_request(&request);
-        for schema_name in catalog_manager.schema_names(&catalog_name).await? {
+        for schema_name in catalog_manager
+            .schema_names(&catalog_name, query_ctx())
+            .await?
+        {
             self.add_namespace(&predicates, &schema_name);
         }
         self.finish()

src/catalog/src/system_schema/utils.rs

@@ -12,47 +12,33 @@
 // See the License for the specific language governing permissions and
 // limitations under the License.
 
-pub mod tables;
+use std::sync::Weak;
 
-use std::sync::{Arc, Weak};
-
-use common_config::Mode;
 use common_meta::key::TableMetadataManagerRef;
-use meta_client::client::MetaClient;
 use snafu::OptionExt;
 
-use crate::error::{Result, UpgradeWeakCatalogManagerRefSnafu};
+use crate::error::{GetInformationExtensionSnafu, Result, UpgradeWeakCatalogManagerRefSnafu};
+use crate::information_schema::InformationExtensionRef;
 use crate::kvbackend::KvBackendCatalogManager;
 use crate::CatalogManager;
 
-/// Try to get the server running mode from `[CatalogManager]` weak reference.
-pub fn running_mode(catalog_manager: &Weak<dyn CatalogManager>) -> Result<Option<Mode>> {
+pub mod tables;
+
+/// Try to get the `[InformationExtension]` from `[CatalogManager]` weak reference.
+pub fn information_extension(
+    catalog_manager: &Weak<dyn CatalogManager>,
+) -> Result<InformationExtensionRef> {
     let catalog_manager = catalog_manager
         .upgrade()
         .context(UpgradeWeakCatalogManagerRefSnafu)?;
 
-    Ok(catalog_manager
+    let information_extension = catalog_manager
         .as_any()
         .downcast_ref::<KvBackendCatalogManager>()
-        .map(|manager| manager.running_mode())
-        .copied())
-}
+        .map(|manager| manager.information_extension())
+        .context(GetInformationExtensionSnafu)?;
 
-/// Try to get the `[MetaClient]` from `[CatalogManager]` weak reference.
-pub fn meta_client(catalog_manager: &Weak<dyn CatalogManager>) -> Result<Option<Arc<MetaClient>>> {
-    let catalog_manager = catalog_manager
-        .upgrade()
-        .context(UpgradeWeakCatalogManagerRefSnafu)?;
-
-    let meta_client = match catalog_manager
-        .as_any()
-        .downcast_ref::<KvBackendCatalogManager>()
-    {
-        None => None,
-        Some(manager) => manager.meta_client(),
-    };
-
-    Ok(meta_client)
+    Ok(information_extension)
 }
 
 /// Try to get the `[TableMetadataManagerRef]` from `[CatalogManager]` weak reference.

src/catalog/src/table_source.rs

@@ -23,7 +23,7 @@ use datafusion::datasource::view::ViewTable;
 use datafusion::datasource::{provider_as_source, TableProvider};
 use datafusion::logical_expr::TableSource;
 use itertools::Itertools;
-use session::context::QueryContext;
+use session::context::QueryContextRef;
 use snafu::{ensure, OptionExt, ResultExt};
 use table::metadata::TableType;
 use table::table::adapter::DfTableProviderAdapter;
@@ -45,6 +45,7 @@ pub struct DfTableSourceProvider {
     disallow_cross_catalog_query: bool,
     default_catalog: String,
     default_schema: String,
+    query_ctx: QueryContextRef,
     plan_decoder: SubstraitPlanDecoderRef,
     enable_ident_normalization: bool,
 }
@@ -53,7 +54,7 @@ impl DfTableSourceProvider {
     pub fn new(
         catalog_manager: CatalogManagerRef,
         disallow_cross_catalog_query: bool,
-        query_ctx: &QueryContext,
+        query_ctx: QueryContextRef,
         plan_decoder: SubstraitPlanDecoderRef,
         enable_ident_normalization: bool,
     ) -> Self {
@@ -63,6 +64,7 @@ impl DfTableSourceProvider {
             resolved_tables: HashMap::new(),
             default_catalog: query_ctx.current_catalog().to_owned(),
             default_schema: query_ctx.current_schema(),
+            query_ctx,
             plan_decoder,
             enable_ident_normalization,
         }
@@ -71,8 +73,7 @@ impl DfTableSourceProvider {
     pub fn resolve_table_ref(&self, table_ref: TableReference) -> Result<ResolvedTableReference> {
         if self.disallow_cross_catalog_query {
             match &table_ref {
-                TableReference::Bare { .. } => (),
-                TableReference::Partial { .. } => {}
+                TableReference::Bare { .. } | TableReference::Partial { .. } => {}
                 TableReference::Full {
                     catalog, schema, ..
                 } => {
@@ -107,7 +108,7 @@ impl DfTableSourceProvider {
 
         let table = self
             .catalog_manager
-            .table(catalog_name, schema_name, table_name)
+            .table(catalog_name, schema_name, table_name, Some(&self.query_ctx))
             .await?
             .with_context(|| TableNotExistSnafu {
                 table: format_full_table_name(catalog_name, schema_name, table_name),
@@ -210,12 +211,12 @@ mod tests {
 
     #[test]
     fn test_validate_table_ref() {
-        let query_ctx = &QueryContext::with("greptime", "public");
+        let query_ctx = Arc::new(QueryContext::with("greptime", "public"));
 
         let table_provider = DfTableSourceProvider::new(
             MemoryCatalogManager::with_default_setup(),
             true,
-            query_ctx,
+            query_ctx.clone(),
             DummyDecoder::arc(),
             true,
         );
@@ -258,7 +259,6 @@ mod tests {
 
     use arrow::datatypes::{DataType, Field, Schema, SchemaRef};
     use cache::{build_fundamental_cache_registry, with_default_composite_cache_registry};
-    use common_config::Mode;
     use common_meta::cache::{CacheRegistryBuilder, LayeredCacheRegistryBuilder};
     use common_meta::key::TableMetadataManager;
     use common_meta::kv_backend::memory::MemoryKvBackend;
@@ -268,6 +268,8 @@ mod tests {
     use datafusion::logical_expr::builder::LogicalTableSource;
     use datafusion::logical_expr::{col, lit, LogicalPlan, LogicalPlanBuilder};
 
+    use crate::information_schema::NoopInformationExtension;
+
     struct MockDecoder;
     impl MockDecoder {
         pub fn arc() -> Arc<Self> {
@@ -308,7 +310,7 @@ mod tests {
 
     #[tokio::test]
     async fn test_resolve_view() {
-        let query_ctx = &QueryContext::with("greptime", "public");
+        let query_ctx = Arc::new(QueryContext::with("greptime", "public"));
         let backend = Arc::new(MemoryKvBackend::default());
         let layered_cache_builder = LayeredCacheRegistryBuilder::default()
             .add_cache_registry(CacheRegistryBuilder::default().build());
@@ -322,10 +324,10 @@ mod tests {
         );
 
         let catalog_manager = KvBackendCatalogManager::new(
-            Mode::Standalone,
-            None,
+            Arc::new(NoopInformationExtension),
             backend.clone(),
             layered_cache_registry,
+            None,
         );
         let table_metadata_manager = TableMetadataManager::new(backend);
         let mut view_info = common_meta::key::test_utils::new_test_table_info(1024, vec![]);
@@ -344,8 +346,13 @@ mod tests {
             .await
             .unwrap();
 
-        let mut table_provider =
-            DfTableSourceProvider::new(catalog_manager, true, query_ctx, MockDecoder::arc(), true);
+        let mut table_provider = DfTableSourceProvider::new(
+            catalog_manager,
+            true,
+            query_ctx.clone(),
+            MockDecoder::arc(),
+            true,
+        );
 
         // View not found
         let table_ref = TableReference::bare("not_exists_view");

src/catalog/src/table_source/dummy_catalog.rs

@@ -112,7 +112,7 @@ impl SchemaProvider for DummySchemaProvider {
     async fn table(&self, name: &str) -> datafusion::error::Result<Option<Arc<dyn TableProvider>>> {
         let table = self
             .catalog_manager
-            .table(&self.catalog_name, &self.schema_name, name)
+            .table(&self.catalog_name, &self.schema_name, name, None)
             .await?
             .with_context(|| TableNotExistSnafu {
                 table: format_full_table_name(&self.catalog_name, &self.schema_name, name),

src/cli/Cargo.toml

@@ -0,0 +1,65 @@
+[package]
+name = "cli"
+version.workspace = true
+edition.workspace = true
+license.workspace = true
+
+[lints]
+workspace = true
+
+[dependencies]
+async-trait.workspace = true
+auth.workspace = true
+base64.workspace = true
+cache.workspace = true
+catalog.workspace = true
+chrono.workspace = true
+clap.workspace = true
+client.workspace = true
+common-base.workspace = true
+common-catalog.workspace = true
+common-config.workspace = true
+common-error.workspace = true
+common-grpc.workspace = true
+common-macro.workspace = true
+common-meta.workspace = true
+common-options.workspace = true
+common-procedure.workspace = true
+common-query.workspace = true
+common-recordbatch.workspace = true
+common-runtime.workspace = true
+common-telemetry = { workspace = true, features = [
+    "deadlock_detection",
+] }
+common-time.workspace = true
+common-version.workspace = true
+common-wal.workspace = true
+datatypes.workspace = true
+either = "1.8"
+etcd-client.workspace = true
+futures.workspace = true
+humantime.workspace = true
+meta-client.workspace = true
+nu-ansi-term = "0.46"
+query.workspace = true
+rand.workspace = true
+reqwest.workspace = true
+rustyline = "10.1"
+serde.workspace = true
+serde_json.workspace = true
+servers.workspace = true
+session.workspace = true
+snafu.workspace = true
+store-api.workspace = true
+substrait.workspace = true
+table.workspace = true
+tokio.workspace = true
+tracing-appender.workspace = true
+
+[dev-dependencies]
+client = { workspace = true, features = ["testing"] }
+common-test-util.workspace = true
+common-version.workspace = true
+serde.workspace = true
+temp-env = "0.3"
+tempfile.workspace = true

src/cli/src/bench.rs

@@ -19,6 +19,7 @@ use std::time::Duration;
 
 use async_trait::async_trait;
 use clap::Parser;
+use common_error::ext::BoxedError;
 use common_meta::key::{TableMetadataManager, TableMetadataManagerRef};
 use common_meta::kv_backend::etcd::EtcdStore;
 use common_meta::peer::Peer;
@@ -30,11 +31,9 @@ use rand::Rng;
 use store_api::storage::RegionNumber;
 use table::metadata::{RawTableInfo, RawTableMeta, TableId, TableIdent, TableType};
 use table::table_name::TableName;
-use tracing_appender::non_blocking::WorkerGuard;
 
 use self::metadata::TableMetadataBencher;
-use crate::cli::{Instance, Tool};
-use crate::error::Result;
+use crate::Tool;
 
 mod metadata;
 
@@ -62,7 +61,7 @@ pub struct BenchTableMetadataCommand {
 }
 
 impl BenchTableMetadataCommand {
-    pub async fn build(&self, guard: Vec<WorkerGuard>) -> Result<Instance> {
+    pub async fn build(&self) -> std::result::Result<Box<dyn Tool>, BoxedError> {
         let etcd_store = EtcdStore::with_endpoints([&self.etcd_addr], 128)
             .await
             .unwrap();
@@ -73,7 +72,7 @@ impl BenchTableMetadataCommand {
             table_metadata_manager,
             count: self.count,
         };
-        Ok(Instance::new(Box::new(tool), guard))
+        Ok(Box::new(tool))
     }
 }
 
@@ -84,7 +83,7 @@ struct BenchTableMetadata {
 
 #[async_trait]
 impl Tool for BenchTableMetadata {
-    async fn do_work(&self) -> Result<()> {
+    async fn do_work(&self) -> std::result::Result<(), BoxedError> {
         let bencher = TableMetadataBencher::new(self.table_metadata_manager.clone(), self.count);
         bencher.bench_create().await;
         bencher.bench_get().await;
@@ -158,7 +157,7 @@ fn create_region_routes(regions: Vec<RegionNumber>) -> Vec<RegionRoute> {
                 addr: String::new(),
             }),
             follower_peers: vec![],
-            leader_status: None,
+            leader_state: None,
             leader_down_since: None,
         });
     }

src/cli/src/bench/metadata.rs

@@ -18,7 +18,7 @@ use common_meta::key::table_route::TableRouteValue;
 use common_meta::key::TableMetadataManagerRef;
 use table::table_name::TableName;
 
-use crate::cli::bench::{
+use crate::bench::{
     bench_self_recorded, create_region_routes, create_region_wal_options, create_table_info,
 };
 

src/cli/src/database.rs

@@ -0,0 +1,136 @@
+// Copyright 2023 Greptime Team
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//     http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+use std::time::Duration;
+
+use base64::engine::general_purpose;
+use base64::Engine;
+use common_catalog::consts::{DEFAULT_CATALOG_NAME, DEFAULT_SCHEMA_NAME};
+use humantime::format_duration;
+use serde_json::Value;
+use servers::http::header::constants::GREPTIME_DB_HEADER_TIMEOUT;
+use servers::http::result::greptime_result_v1::GreptimedbV1Response;
+use servers::http::GreptimeQueryOutput;
+use snafu::ResultExt;
+
+use crate::error::{HttpQuerySqlSnafu, Result, SerdeJsonSnafu};
+
+#[derive(Debug, Clone)]
+pub struct DatabaseClient {
+    addr: String,
+    catalog: String,
+    auth_header: Option<String>,
+    timeout: Duration,
+}
+
+impl DatabaseClient {
+    pub fn new(
+        addr: String,
+        catalog: String,
+        auth_basic: Option<String>,
+        timeout: Duration,
+    ) -> Self {
+        let auth_header = if let Some(basic) = auth_basic {
+            let encoded = general_purpose::STANDARD.encode(basic);
+            Some(format!("basic {}", encoded))
+        } else {
+            None
+        };
+
+        Self {
+            addr,
+            catalog,
+            auth_header,
+            timeout,
+        }
+    }
+
+    pub async fn sql_in_public(&self, sql: &str) -> Result<Option<Vec<Vec<Value>>>> {
+        self.sql(sql, DEFAULT_SCHEMA_NAME).await
+    }
+
+    /// Execute sql query.
+    pub async fn sql(&self, sql: &str, schema: &str) -> Result<Option<Vec<Vec<Value>>>> {
+        let url = format!("http://{}/v1/sql", self.addr);
+        let params = [
+            ("db", format!("{}-{}", self.catalog, schema)),
+            ("sql", sql.to_string()),
+        ];
+        let mut request = reqwest::Client::new()
+            .post(&url)
+            .form(&params)
+            .header("Content-Type", "application/x-www-form-urlencoded");
+        if let Some(ref auth) = self.auth_header {
+            request = request.header("Authorization", auth);
+        }
+
+        request = request.header(
+            GREPTIME_DB_HEADER_TIMEOUT,
+            format_duration(self.timeout).to_string(),
+        );
+
+        let response = request.send().await.with_context(|_| HttpQuerySqlSnafu {
+            reason: format!("bad url: {}", url),
+        })?;
+        let response = response
+            .error_for_status()
+            .with_context(|_| HttpQuerySqlSnafu {
+                reason: format!("query failed: {}", sql),
+            })?;
+
+        let text = response.text().await.with_context(|_| HttpQuerySqlSnafu {
+            reason: "cannot get response text".to_string(),
+        })?;
+
+        let body = serde_json::from_str::<GreptimedbV1Response>(&text).context(SerdeJsonSnafu)?;
+        Ok(body.output().first().and_then(|output| match output {
+            GreptimeQueryOutput::Records(records) => Some(records.rows().clone()),
+            GreptimeQueryOutput::AffectedRows(_) => None,
+        }))
+    }
+}
+
+/// Split at `-`.
+pub(crate) fn split_database(database: &str) -> Result<(String, Option<String>)> {
+    let (catalog, schema) = match database.split_once('-') {
+        Some((catalog, schema)) => (catalog, schema),
+        None => (DEFAULT_CATALOG_NAME, database),
+    };
+
+    if schema == "*" {
+        Ok((catalog.to_string(), None))
+    } else {
+        Ok((catalog.to_string(), Some(schema.to_string())))
+    }
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    #[test]
+    fn test_split_database() {
+        let result = split_database("catalog-schema").unwrap();
+        assert_eq!(result, ("catalog".to_string(), Some("schema".to_string())));
+
+        let result = split_database("schema").unwrap();
+        assert_eq!(result, ("greptime".to_string(), Some("schema".to_string())));
+
+        let result = split_database("catalog-*").unwrap();
+        assert_eq!(result, ("catalog".to_string(), None));
+
+        let result = split_database("*").unwrap();
+        assert_eq!(result, ("greptime".to_string(), None));
+    }
+}

src/cli/src/error.rs

@@ -0,0 +1,316 @@
+// Copyright 2023 Greptime Team
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//     http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+use std::any::Any;
+
+use common_error::ext::{BoxedError, ErrorExt};
+use common_error::status_code::StatusCode;
+use common_macro::stack_trace_debug;
+use rustyline::error::ReadlineError;
+use snafu::{Location, Snafu};
+
+#[derive(Snafu)]
+#[snafu(visibility(pub))]
+#[stack_trace_debug]
+pub enum Error {
+    #[snafu(display("Failed to install ring crypto provider: {}", msg))]
+    InitTlsProvider {
+        #[snafu(implicit)]
+        location: Location,
+        msg: String,
+    },
+    #[snafu(display("Failed to create default catalog and schema"))]
+    InitMetadata {
+        #[snafu(implicit)]
+        location: Location,
+        source: common_meta::error::Error,
+    },
+
+    #[snafu(display("Failed to init DDL manager"))]
+    InitDdlManager {
+        #[snafu(implicit)]
+        location: Location,
+        source: common_meta::error::Error,
+    },
+
+    #[snafu(display("Failed to init default timezone"))]
+    InitTimezone {
+        #[snafu(implicit)]
+        location: Location,
+        source: common_time::error::Error,
+    },
+
+    #[snafu(display("Failed to start procedure manager"))]
+    StartProcedureManager {
+        #[snafu(implicit)]
+        location: Location,
+        source: common_procedure::error::Error,
+    },
+
+    #[snafu(display("Failed to stop procedure manager"))]
+    StopProcedureManager {
+        #[snafu(implicit)]
+        location: Location,
+        source: common_procedure::error::Error,
+    },
+
+    #[snafu(display("Failed to start wal options allocator"))]
+    StartWalOptionsAllocator {
+        #[snafu(implicit)]
+        location: Location,
+        source: common_meta::error::Error,
+    },
+
+    #[snafu(display("Missing config, msg: {}", msg))]
+    MissingConfig {
+        msg: String,
+        #[snafu(implicit)]
+        location: Location,
+    },
+
+    #[snafu(display("Illegal config: {}", msg))]
+    IllegalConfig {
+        msg: String,
+        #[snafu(implicit)]
+        location: Location,
+    },
+
+    #[snafu(display("Invalid REPL command: {reason}"))]
+    InvalidReplCommand { reason: String },
+
+    #[snafu(display("Cannot create REPL"))]
+    ReplCreation {
+        #[snafu(source)]
+        error: ReadlineError,
+        #[snafu(implicit)]
+        location: Location,
+    },
+
+    #[snafu(display("Error reading command"))]
+    Readline {
+        #[snafu(source)]
+        error: ReadlineError,
+        #[snafu(implicit)]
+        location: Location,
+    },
+
+    #[snafu(display("Failed to request database, sql: {sql}"))]
+    RequestDatabase {
+        sql: String,
+        #[snafu(source)]
+        source: client::Error,
+        #[snafu(implicit)]
+        location: Location,
+    },
+
+    #[snafu(display("Failed to collect RecordBatches"))]
+    CollectRecordBatches {
+        #[snafu(implicit)]
+        location: Location,
+        source: common_recordbatch::error::Error,
+    },
+
+    #[snafu(display("Failed to pretty print Recordbatches"))]
+    PrettyPrintRecordBatches {
+        #[snafu(implicit)]
+        location: Location,
+        source: common_recordbatch::error::Error,
+    },
+
+    #[snafu(display("Failed to start Meta client"))]
+    StartMetaClient {
+        #[snafu(implicit)]
+        location: Location,
+        source: meta_client::error::Error,
+    },
+
+    #[snafu(display("Failed to parse SQL: {}", sql))]
+    ParseSql {
+        sql: String,
+        #[snafu(implicit)]
+        location: Location,
+        source: query::error::Error,
+    },
+
+    #[snafu(display("Failed to plan statement"))]
+    PlanStatement {
+        #[snafu(implicit)]
+        location: Location,
+        source: query::error::Error,
+    },
+
+    #[snafu(display("Failed to encode logical plan in substrait"))]
+    SubstraitEncodeLogicalPlan {
+        #[snafu(implicit)]
+        location: Location,
+        source: substrait::error::Error,
+    },
+
+    #[snafu(display("Failed to load layered config"))]
+    LoadLayeredConfig {
+        #[snafu(source(from(common_config::error::Error, Box::new)))]
+        source: Box<common_config::error::Error>,
+        #[snafu(implicit)]
+        location: Location,
+    },
+
+    #[snafu(display("Failed to connect to Etcd at {etcd_addr}"))]
+    ConnectEtcd {
+        etcd_addr: String,
+        #[snafu(source)]
+        error: etcd_client::Error,
+        #[snafu(implicit)]
+        location: Location,
+    },
+
+    #[snafu(display("Failed to serde json"))]
+    SerdeJson {
+        #[snafu(source)]
+        error: serde_json::error::Error,
+        #[snafu(implicit)]
+        location: Location,
+    },
+
+    #[snafu(display("Failed to run http request: {reason}"))]
+    HttpQuerySql {
+        reason: String,
+        #[snafu(source)]
+        error: reqwest::Error,
+        #[snafu(implicit)]
+        location: Location,
+    },
+
+    #[snafu(display("Empty result from output"))]
+    EmptyResult {
+        #[snafu(implicit)]
+        location: Location,
+    },
+
+    #[snafu(display("Failed to manipulate file"))]
+    FileIo {
+        #[snafu(implicit)]
+        location: Location,
+        #[snafu(source)]
+        error: std::io::Error,
+    },
+
+    #[snafu(display("Failed to create directory {}", dir))]
+    CreateDir {
+        dir: String,
+        #[snafu(source)]
+        error: std::io::Error,
+    },
+
+    #[snafu(display("Failed to spawn thread"))]
+    SpawnThread {
+        #[snafu(source)]
+        error: std::io::Error,
+    },
+
+    #[snafu(display("Other error"))]
+    Other {
+        source: BoxedError,
+        #[snafu(implicit)]
+        location: Location,
+    },
+
+    #[snafu(display("Failed to build runtime"))]
+    BuildRuntime {
+        #[snafu(implicit)]
+        location: Location,
+        source: common_runtime::error::Error,
+    },
+
+    #[snafu(display("Failed to get cache from cache registry: {}", name))]
+    CacheRequired {
+        #[snafu(implicit)]
+        location: Location,
+        name: String,
+    },
+
+    #[snafu(display("Failed to build cache registry"))]
+    BuildCacheRegistry {
+        #[snafu(implicit)]
+        location: Location,
+        source: cache::error::Error,
+    },
+
+    #[snafu(display("Failed to initialize meta client"))]
+    MetaClientInit {
+        #[snafu(implicit)]
+        location: Location,
+        source: meta_client::error::Error,
+    },
+
+    #[snafu(display("Cannot find schema {schema} in catalog {catalog}"))]
+    SchemaNotFound {
+        catalog: String,
+        schema: String,
+        #[snafu(implicit)]
+        location: Location,
+    },
+}
+
+pub type Result<T> = std::result::Result<T, Error>;
+
+impl ErrorExt for Error {
+    fn status_code(&self) -> StatusCode {
+        match self {
+            Error::InitMetadata { source, .. } | Error::InitDdlManager { source, .. } => {
+                source.status_code()
+            }
+
+            Error::MissingConfig { .. }
+            | Error::LoadLayeredConfig { .. }
+            | Error::IllegalConfig { .. }
+            | Error::InvalidReplCommand { .. }
+            | Error::InitTimezone { .. }
+            | Error::ConnectEtcd { .. }
+            | Error::CreateDir { .. }
+            | Error::EmptyResult { .. } => StatusCode::InvalidArguments,
+
+            Error::StartProcedureManager { source, .. }
+            | Error::StopProcedureManager { source, .. } => source.status_code(),
+            Error::StartWalOptionsAllocator { source, .. } => source.status_code(),
+            Error::ReplCreation { .. } | Error::Readline { .. } | Error::HttpQuerySql { .. } => {
+                StatusCode::Internal
+            }
+            Error::RequestDatabase { source, .. } => source.status_code(),
+            Error::CollectRecordBatches { source, .. }
+            | Error::PrettyPrintRecordBatches { source, .. } => source.status_code(),
+            Error::StartMetaClient { source, .. } => source.status_code(),
+            Error::ParseSql { source, .. } | Error::PlanStatement { source, .. } => {
+                source.status_code()
+            }
+            Error::SubstraitEncodeLogicalPlan { source, .. } => source.status_code(),
+
+            Error::SerdeJson { .. }
+            | Error::FileIo { .. }
+            | Error::SpawnThread { .. }
+            | Error::InitTlsProvider { .. } => StatusCode::Unexpected,
+
+            Error::Other { source, .. } => source.status_code(),
+
+            Error::BuildRuntime { source, .. } => source.status_code(),
+
+            Error::CacheRequired { .. } | Error::BuildCacheRegistry { .. } => StatusCode::Internal,
+            Error::MetaClientInit { source, .. } => source.status_code(),
+            Error::SchemaNotFound { .. } => StatusCode::DatabaseNotFound,
+        }
+    }
+
+    fn as_any(&self) -> &dyn Any {
+        self
+    }
+}

src/cli/src/export.rs

@@ -0,0 +1,484 @@
+// Copyright 2023 Greptime Team
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//     http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+use std::collections::HashSet;
+use std::path::PathBuf;
+use std::sync::Arc;
+use std::time::Duration;
+
+use async_trait::async_trait;
+use clap::{Parser, ValueEnum};
+use common_error::ext::BoxedError;
+use common_telemetry::{debug, error, info};
+use serde_json::Value;
+use snafu::{OptionExt, ResultExt};
+use tokio::fs::File;
+use tokio::io::{AsyncWriteExt, BufWriter};
+use tokio::sync::Semaphore;
+use tokio::time::Instant;
+
+use crate::database::DatabaseClient;
+use crate::error::{EmptyResultSnafu, Error, FileIoSnafu, Result, SchemaNotFoundSnafu};
+use crate::{database, Tool};
+
+type TableReference = (String, String, String);
+
+#[derive(Debug, Default, Clone, ValueEnum)]
+enum ExportTarget {
+    /// Export all table schemas, corresponding to `SHOW CREATE TABLE`.
+    Schema,
+    /// Export all table data, corresponding to `COPY DATABASE TO`.
+    Data,
+    /// Export all table schemas and data at once.
+    #[default]
+    All,
+}
+
+#[derive(Debug, Default, Parser)]
+pub struct ExportCommand {
+    /// Server address to connect
+    #[clap(long)]
+    addr: String,
+
+    /// Directory to put the exported data. E.g.: /tmp/greptimedb-export
+    #[clap(long)]
+    output_dir: String,
+
+    /// The name of the catalog to export.
+    #[clap(long, default_value = "greptime-*")]
+    database: String,
+
+    /// Parallelism of the export.
+    #[clap(long, short = 'j', default_value = "1")]
+    export_jobs: usize,
+
+    /// Max retry times for each job.
+    #[clap(long, default_value = "3")]
+    max_retry: usize,
+
+    /// Things to export
+    #[clap(long, short = 't', value_enum, default_value = "all")]
+    target: ExportTarget,
+
+    /// A half-open time range: [start_time, end_time).
+    /// The start of the time range (time-index column) for data export.
+    #[clap(long)]
+    start_time: Option<String>,
+
+    /// A half-open time range: [start_time, end_time).
+    /// The end of the time range (time-index column) for data export.
+    #[clap(long)]
+    end_time: Option<String>,
+
+    /// The basic authentication for connecting to the server
+    #[clap(long)]
+    auth_basic: Option<String>,
+
+    /// The timeout of invoking the database.
+    ///
+    /// It is used to override the server-side timeout setting.
+    /// The default behavior will disable server-side default timeout(i.e. `0s`).
+    #[clap(long, value_parser = humantime::parse_duration)]
+    timeout: Option<Duration>,
+}
+
+impl ExportCommand {
+    pub async fn build(&self) -> std::result::Result<Box<dyn Tool>, BoxedError> {
+        let (catalog, schema) =
+            database::split_database(&self.database).map_err(BoxedError::new)?;
+
+        let database_client = DatabaseClient::new(
+            self.addr.clone(),
+            catalog.clone(),
+            self.auth_basic.clone(),
+            // Treats `None` as `0s` to disable server-side default timeout.
+            self.timeout.unwrap_or_default(),
+        );
+
+        Ok(Box::new(Export {
+            catalog,
+            schema,
+            database_client,
+            output_dir: self.output_dir.clone(),
+            parallelism: self.export_jobs,
+            target: self.target.clone(),
+            start_time: self.start_time.clone(),
+            end_time: self.end_time.clone(),
+        }))
+    }
+}
+
+pub struct Export {
+    catalog: String,
+    schema: Option<String>,
+    database_client: DatabaseClient,
+    output_dir: String,
+    parallelism: usize,
+    target: ExportTarget,
+    start_time: Option<String>,
+    end_time: Option<String>,
+}
+
+impl Export {
+    fn catalog_path(&self) -> PathBuf {
+        PathBuf::from(&self.output_dir).join(&self.catalog)
+    }
+
+    async fn get_db_names(&self) -> Result<Vec<String>> {
+        let db_names = self.all_db_names().await?;
+        let Some(schema) = &self.schema else {
+            return Ok(db_names);
+        };
+
+        // Check if the schema exists
+        db_names
+            .into_iter()
+            .find(|db_name| db_name.to_lowercase() == schema.to_lowercase())
+            .map(|name| vec![name])
+            .context(SchemaNotFoundSnafu {
+                catalog: &self.catalog,
+                schema,
+            })
+    }
+
+    /// Iterate over all db names.
+    async fn all_db_names(&self) -> Result<Vec<String>> {
+        let records = self
+            .database_client
+            .sql_in_public("SHOW DATABASES")
+            .await?
+            .context(EmptyResultSnafu)?;
+        let mut result = Vec::with_capacity(records.len());
+        for value in records {
+            let Value::String(schema) = &value[0] else {
+                unreachable!()
+            };
+            if schema == common_catalog::consts::INFORMATION_SCHEMA_NAME {
+                continue;
+            }
+            if schema == common_catalog::consts::PG_CATALOG_NAME {
+                continue;
+            }
+            result.push(schema.clone());
+        }
+        Ok(result)
+    }
+
+    /// Return a list of [`TableReference`] to be exported.
+    /// Includes all tables under the given `catalog` and `schema`.
+    async fn get_table_list(
+        &self,
+        catalog: &str,
+        schema: &str,
+    ) -> Result<(
+        Vec<TableReference>,
+        Vec<TableReference>,
+        Vec<TableReference>,
+    )> {
+        // Puts all metric table first
+        let sql = format!(
+            "SELECT table_catalog, table_schema, table_name \
+            FROM information_schema.columns \
+            WHERE column_name = '__tsid' \
+                and table_catalog = \'{catalog}\' \
+                and table_schema = \'{schema}\'"
+        );
+        let records = self
+            .database_client
+            .sql_in_public(&sql)
+            .await?
+            .context(EmptyResultSnafu)?;
+        let mut metric_physical_tables = HashSet::with_capacity(records.len());
+        for value in records {
+            let mut t = Vec::with_capacity(3);
+            for v in &value {
+                let Value::String(value) = v else {
+                    unreachable!()
+                };
+                t.push(value);
+            }
+            metric_physical_tables.insert((t[0].clone(), t[1].clone(), t[2].clone()));
+        }
+
+        let sql = format!(
+            "SELECT table_catalog, table_schema, table_name, table_type \
+            FROM information_schema.tables \
+            WHERE (table_type = \'BASE TABLE\' OR table_type = \'VIEW\') \
+                and table_catalog = \'{catalog}\' \
+                and table_schema = \'{schema}\'",
+        );
+        let records = self
+            .database_client
+            .sql_in_public(&sql)
+            .await?
+            .context(EmptyResultSnafu)?;
+
+        debug!("Fetched table/view list: {:?}", records);
+
+        if records.is_empty() {
+            return Ok((vec![], vec![], vec![]));
+        }
+
+        let mut remaining_tables = Vec::with_capacity(records.len());
+        let mut views = Vec::new();
+        for value in records {
+            let mut t = Vec::with_capacity(4);
+            for v in &value {
+                let Value::String(value) = v else {
+                    unreachable!()
+                };
+                t.push(value);
+            }
+            let table = (t[0].clone(), t[1].clone(), t[2].clone());
+            let table_type = t[3].as_str();
+            // Ignores the physical table
+            if !metric_physical_tables.contains(&table) {
+                if table_type == "VIEW" {
+                    views.push(table);
+                } else {
+                    remaining_tables.push(table);
+                }
+            }
+        }
+
+        Ok((
+            metric_physical_tables.into_iter().collect(),
+            remaining_tables,
+            views,
+        ))
+    }
+
+    async fn show_create(
+        &self,
+        show_type: &str,
+        catalog: &str,
+        schema: &str,
+        table: Option<&str>,
+    ) -> Result<String> {
+        let sql = match table {
+            Some(table) => format!(
+                r#"SHOW CREATE {} "{}"."{}"."{}""#,
+                show_type, catalog, schema, table
+            ),
+            None => format!(r#"SHOW CREATE {} "{}"."{}""#, show_type, catalog, schema),
+        };
+        let records = self
+            .database_client
+            .sql_in_public(&sql)
+            .await?
+            .context(EmptyResultSnafu)?;
+        let Value::String(create) = &records[0][1] else {
+            unreachable!()
+        };
+
+        Ok(format!("{};\n", create))
+    }
+
+    async fn export_create_database(&self) -> Result<()> {
+        let timer = Instant::now();
+        let db_names = self.get_db_names().await?;
+        let db_count = db_names.len();
+        for schema in db_names {
+            let db_dir = self.catalog_path().join(format!("{schema}/"));
+            tokio::fs::create_dir_all(&db_dir)
+                .await
+                .context(FileIoSnafu)?;
+            let file = db_dir.join("create_database.sql");
+            let mut file = File::create(file).await.context(FileIoSnafu)?;
+            let create_database = self
+                .show_create("DATABASE", &self.catalog, &schema, None)
+                .await?;
+            file.write_all(create_database.as_bytes())
+                .await
+                .context(FileIoSnafu)?;
+        }
+
+        let elapsed = timer.elapsed();
+        info!("Success {db_count} jobs, cost: {elapsed:?}");
+
+        Ok(())
+    }
+
+    async fn export_create_table(&self) -> Result<()> {
+        let timer = Instant::now();
+        let semaphore = Arc::new(Semaphore::new(self.parallelism));
+        let db_names = self.get_db_names().await?;
+        let db_count = db_names.len();
+        let mut tasks = Vec::with_capacity(db_names.len());
+        for schema in db_names {
+            let semaphore_moved = semaphore.clone();
+            tasks.push(async move {
+                let _permit = semaphore_moved.acquire().await.unwrap();
+                let (metric_physical_tables, remaining_tables, views) =
+                    self.get_table_list(&self.catalog, &schema).await?;
+                let table_count =
+                    metric_physical_tables.len() + remaining_tables.len() + views.len();
+                let db_dir = self.catalog_path().join(format!("{schema}/"));
+                tokio::fs::create_dir_all(&db_dir)
+                    .await
+                    .context(FileIoSnafu)?;
+                let file = db_dir.join("create_tables.sql");
+                let mut file = File::create(file).await.context(FileIoSnafu)?;
+                for (c, s, t) in metric_physical_tables.into_iter().chain(remaining_tables) {
+                    let create_table = self.show_create("TABLE", &c, &s, Some(&t)).await?;
+                    file.write_all(create_table.as_bytes())
+                        .await
+                        .context(FileIoSnafu)?;
+                }
+                for (c, s, v) in views {
+                    let create_view = self.show_create("VIEW", &c, &s, Some(&v)).await?;
+                    file.write_all(create_view.as_bytes())
+                        .await
+                        .context(FileIoSnafu)?;
+                }
+
+                info!(
+                    "Finished exporting {}.{schema} with {table_count} table schemas to path: {}",
+                    self.catalog,
+                    db_dir.to_string_lossy()
+                );
+
+                Ok::<(), Error>(())
+            });
+        }
+
+        let success = futures::future::join_all(tasks)
+            .await
+            .into_iter()
+            .filter(|r| match r {
+                Ok(_) => true,
+                Err(e) => {
+                    error!(e; "export schema job failed");
+                    false
+                }
+            })
+            .count();
+
+        let elapsed = timer.elapsed();
+        info!("Success {success}/{db_count} jobs, cost: {elapsed:?}");
+
+        Ok(())
+    }
+
+    async fn export_database_data(&self) -> Result<()> {
+        let timer = Instant::now();
+        let semaphore = Arc::new(Semaphore::new(self.parallelism));
+        let db_names = self.get_db_names().await?;
+        let db_count = db_names.len();
+        let mut tasks = Vec::with_capacity(db_count);
+        for schema in db_names {
+            let semaphore_moved = semaphore.clone();
+            tasks.push(async move {
+                let _permit = semaphore_moved.acquire().await.unwrap();
+                let db_dir = self.catalog_path().join(format!("{schema}/"));
+                tokio::fs::create_dir_all(&db_dir)
+                    .await
+                    .context(FileIoSnafu)?;
+
+                let with_options = match (&self.start_time, &self.end_time) {
+                    (Some(start_time), Some(end_time)) => {
+                        format!(
+                            "WITH (FORMAT='parquet', start_time='{}', end_time='{}')",
+                            start_time, end_time
+                        )
+                    }
+                    (Some(start_time), None) => {
+                        format!("WITH (FORMAT='parquet', start_time='{}')", start_time)
+                    }
+                    (None, Some(end_time)) => {
+                        format!("WITH (FORMAT='parquet', end_time='{}')", end_time)
+                    }
+                    (None, None) => "WITH (FORMAT='parquet')".to_string(),
+                };
+
+                let sql = format!(
+                    r#"COPY DATABASE "{}"."{}" TO '{}' {};"#,
+                    self.catalog,
+                    schema,
+                    db_dir.to_str().unwrap(),
+                    with_options
+                );
+
+                info!("Executing sql: {sql}");
+
+                self.database_client.sql_in_public(&sql).await?;
+
+                info!(
+                    "Finished exporting {}.{schema} data into path: {}",
+                    self.catalog,
+                    db_dir.to_string_lossy()
+                );
+
+                // The export copy from sql
+                let copy_from_file = db_dir.join("copy_from.sql");
+                let mut writer =
+                    BufWriter::new(File::create(copy_from_file).await.context(FileIoSnafu)?);
+                let copy_database_from_sql = format!(
+                    r#"COPY DATABASE "{}"."{}" FROM '{}' WITH (FORMAT='parquet');"#,
+                    self.catalog,
+                    schema,
+                    db_dir.to_str().unwrap()
+                );
+                writer
+                    .write(copy_database_from_sql.as_bytes())
+                    .await
+                    .context(FileIoSnafu)?;
+                writer.flush().await.context(FileIoSnafu)?;
+
+                info!("Finished exporting {}.{schema} copy_from.sql", self.catalog);
+
+                Ok::<(), Error>(())
+            })
+        }
+
+        let success = futures::future::join_all(tasks)
+            .await
+            .into_iter()
+            .filter(|r| match r {
+                Ok(_) => true,
+                Err(e) => {
+                    error!(e; "export database job failed");
+                    false
+                }
+            })
+            .count();
+        let elapsed = timer.elapsed();
+
+        info!("Success {success}/{db_count} jobs, costs: {elapsed:?}");
+
+        Ok(())
+    }
+}
+
+#[async_trait]
+impl Tool for Export {
+    async fn do_work(&self) -> std::result::Result<(), BoxedError> {
+        match self.target {
+            ExportTarget::Schema => {
+                self.export_create_database()
+                    .await
+                    .map_err(BoxedError::new)?;
+                self.export_create_table().await.map_err(BoxedError::new)
+            }
+            ExportTarget::Data => self.export_database_data().await.map_err(BoxedError::new),
+            ExportTarget::All => {
+                self.export_create_database()
+                    .await
+                    .map_err(BoxedError::new)?;
+                self.export_create_table().await.map_err(BoxedError::new)?;
+                self.export_database_data().await.map_err(BoxedError::new)
+            }
+        }
+    }
+}