ci: upload .pdb files too

* chore: add log for converting region to follower

* chore: apply suggestions from CR

.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-greptime-binary/action.yml

@@ -54,7 +54,7 @@ runs:
         PROFILE_TARGET: ${{ inputs.cargo-profile == 'dev' && 'debug' || inputs.cargo-profile }}
       with:
         artifacts-dir: ${{ inputs.artifacts-dir }}
-        target-file: ./target/$PROFILE_TARGET/greptime
+        target-files: ./target/$PROFILE_TARGET/greptime
         version: ${{ inputs.version }}
         working-dir: ${{ inputs.working-dir }}
 
@@ -72,6 +72,6 @@ runs:
       if: ${{ inputs.build-android-artifacts == 'true' }}
       with:
         artifacts-dir: ${{ inputs.artifacts-dir }}
-        target-file: ./target/aarch64-linux-android/release/greptime
+        target-files: ./target/aarch64-linux-android/release/greptime
         version: ${{ inputs.version }}
         working-dir: ${{ inputs.working-dir }}

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

@@ -90,5 +90,5 @@ runs:
       uses: ./.github/actions/upload-artifacts
       with:
         artifacts-dir: ${{ inputs.artifacts-dir }}
-        target-file: target/${{ inputs.arch }}/${{ inputs.cargo-profile }}/greptime
+        target-files: target/${{ inputs.arch }}/${{ inputs.cargo-profile }}/greptime
         version: ${{ inputs.version }}

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

@@ -40,7 +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
@@ -76,5 +76,5 @@ runs:
       uses: ./.github/actions/upload-artifacts
       with:
         artifacts-dir: ${{ inputs.artifacts-dir }}
-        target-file: target/${{ inputs.arch }}/${{ inputs.cargo-profile }}/greptime
+        target-files: target/${{ inputs.arch }}/${{ inputs.cargo-profile }}/greptime,target/${{ inputs.arch }}/${{ inputs.cargo-profile }}/greptime.pdb
         version: ${{ inputs.version }}

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

@@ -18,7 +18,7 @@ runs:
         --set replicaCount=${{ inputs.etcd-replicas }} \
         --set resources.requests.cpu=50m \
         --set resources.requests.memory=128Mi \
-        --set resources.limits.cpu=1000m \
+        --set resources.limits.cpu=1500m \
         --set resources.limits.memory=2Gi \
         --set auth.rbac.create=false \
         --set auth.rbac.token.enabled=false \

.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-greptimedb-cluster/with-minio-and-cache.yaml

@@ -5,7 +5,7 @@ meta:
 
     [datanode]
     [datanode.client]
-    timeout = "60s"
+    timeout = "120s"
 datanode:
   configData: |-
     [runtime]
@@ -21,7 +21,7 @@ frontend:
     global_rt_size = 4
 
     [meta_client]
-    ddl_timeout = "60s"
+    ddl_timeout = "120s"
 objectStorage:
   s3:
     bucket: default

.github/actions/setup-greptimedb-cluster/with-minio.yaml

@@ -5,7 +5,7 @@ meta:
     
     [datanode]
     [datanode.client]
-    timeout = "60s"
+    timeout = "120s"
 datanode:
   configData: |-
     [runtime]
@@ -17,7 +17,7 @@ frontend:
     global_rt_size = 4
 
     [meta_client]
-    ddl_timeout = "60s"
+    ddl_timeout = "120s"
 objectStorage:
   s3:
     bucket: default

.github/actions/setup-greptimedb-cluster/with-remote-wal.yaml

@@ -11,7 +11,7 @@ meta:
         
     [datanode]
     [datanode.client]
-    timeout = "60s"
+    timeout = "120s"
 datanode:
   configData: |-
     [runtime]
@@ -28,7 +28,7 @@ frontend:
     global_rt_size = 4
 
     [meta_client]
-    ddl_timeout = "60s"
+    ddl_timeout = "120s"
 objectStorage:
   s3:
     bucket: default

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

@@ -18,6 +18,8 @@ runs:
         --set controller.replicaCount=${{ inputs.controller-replicas }} \
         --set controller.resources.requests.cpu=50m \
         --set controller.resources.requests.memory=128Mi \
+        --set controller.resources.limits.cpu=2000m \
+        --set controller.resources.limits.memory=2Gi \
         --set listeners.controller.protocol=PLAINTEXT \
         --set listeners.client.protocol=PLAINTEXT \
         --create-namespace \

.github/actions/upload-artifacts/action.yml

@@ -4,8 +4,8 @@ inputs:
   artifacts-dir:
     description: Directory to store artifacts
     required: true
-  target-file:
-    description: The path of the target artifact
+  target-files:
+    description: The multiple target files to upload, separated by comma
     required: false
   version:
     description: Version of the artifact
@@ -18,12 +18,16 @@ runs:
   using: composite
   steps:
     - name: Create artifacts directory
-      if: ${{ inputs.target-file != '' }}
+      if: ${{ inputs.target-files != '' }}
       working-directory: ${{ inputs.working-dir }}
       shell: bash
       run: |
-        mkdir -p ${{ inputs.artifacts-dir }} && \
-        cp ${{ inputs.target-file }} ${{ inputs.artifacts-dir }}
+        set -e
+        mkdir -p ${{ inputs.artifacts-dir }}
+        IFS=',' read -ra FILES <<< "${{ inputs.target-files }}"
+        for file in "${FILES[@]}"; do
+          cp "$file" ${{ inputs.artifacts-dir }}/
+        done
 
     # The compressed artifacts will use the following layout:
     # greptime-linux-amd64-pyo3-v0.3.0sha256sum

.github/cargo-blacklist.txt

@@ -0,0 +1,3 @@
+native-tls
+openssl
+aws-lc-sys

.github/pull_request_template.md

@@ -4,7 +4,8 @@ I hereby agree to the terms of the [GreptimeDB CLA](https://github.com/GreptimeT
 
 ## What's changed and what's your intention?
 
-__!!! DO NOT LEAVE THIS BLOCK EMPTY !!!__
+<!--    
+ __!!! DO NOT LEAVE THIS BLOCK EMPTY !!!__
 
 Please explain IN DETAIL what the changes are in this PR and why they are needed:
 
@@ -12,9 +13,14 @@ Please explain IN DETAIL what the changes are in this PR and why they are needed
 - How does this PR work? Need a brief introduction for the changed logic (optional)
 - Describe clearly one logical change and avoid lazy messages (optional)
 - Describe any limitations of the current code (optional)
+- Describe if this PR will break **API or data compatibility**  (optional)
+-->
 
-## Checklist
+## PR Checklist
+Please convert it to a draft if some of the following conditions are not met.
 
 - [ ] I have written the necessary rustdoc comments.
 - [ ] I have added the necessary unit tests and integration tests.
 - [ ] This PR requires documentation updates.
+- [ ] API changes are backward compatible.
+- [ ] Schema or data changes are backward compatible.

.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/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/develop.yml

@@ -323,8 +323,6 @@ 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:
@@ -429,12 +427,25 @@ jobs:
     timeout-minutes: 60
     strategy:
       matrix:
-        target: ["fuzz_migrate_mito_regions", "fuzz_failover_mito_regions", "fuzz_failover_metric_regions"]
+        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: |
@@ -461,8 +472,6 @@ 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:
@@ -523,7 +532,7 @@ 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&
@@ -674,7 +683,7 @@ 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

.github/workflows/nightly-build.yml

@@ -12,7 +12,7 @@ on:
       linux_amd64_runner:
         type: choice
         description: The runner uses to build linux-amd64 artifacts
-        default: ec2-c6i.2xlarge-amd64
+        default: ec2-c6i.4xlarge-amd64
         options:
           - ubuntu-20.04
           - ubuntu-20.04-8-cores
@@ -27,7 +27,7 @@ on:
       linux_arm64_runner:
         type: choice
         description: The runner uses to build linux-arm64 artifacts
-        default: ec2-c6g.2xlarge-arm64
+        default: ec2-c6g.4xlarge-arm64
         options:
           - ec2-c6g.xlarge-arm64 # 4C8G
           - ec2-c6g.2xlarge-arm64 # 8C16G

.github/workflows/nightly-ci.yml

@@ -22,6 +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:
@@ -92,7 +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:
@@ -110,6 +114,16 @@ jobs:
           GT_S3_REGION: ${{ vars.AWS_CI_TEST_BUCKET_REGION }}
           UNITTEST_LOG_DIR: "__unittest_logs"
 
+  cleanbuild-linux-nix:
+    runs-on: ubuntu-latest-8-cores
+    timeout-minutes: 60
+    steps:
+      - uses: actions/checkout@v4
+      - uses: cachix/install-nix-action@v27
+        with:
+          nix_path: nixpkgs=channel:nixos-unstable
+      - run: nix-shell --pure --run "cargo build"
+
   check-status:
     name: Check status
     needs: [sqlness-test, sqlness-windows, test-on-windows]

.github/workflows/release.yml

@@ -31,7 +31,7 @@ on:
       linux_arm64_runner:
         type: choice
         description: The runner uses to build linux-arm64 artifacts
-        default: ec2-c6g.4xlarge-arm64
+        default: ec2-c6g.8xlarge-arm64
         options:
           - ubuntu-2204-32-cores-arm
           - ec2-c6g.xlarge-arm64 # 4C8G
@@ -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: v0.12.0
 
 # Permission reference: https://docs.github.com/en/actions/using-jobs/assigning-permissions-to-jobs
 permissions:

.gitignore

@@ -47,6 +47,10 @@ benchmarks/data
 
 venv/
 
-# Fuzz tests 
+# Fuzz tests
 tests-fuzz/artifacts/
 tests-fuzz/corpus/
+
+# Nix
+.direnv
+.envrc

.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 = "0.9.3"
+version = "0.12.0"
 edition = "2021"
 license = "Apache-2.0"
 
@@ -90,7 +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"
@@ -99,7 +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"
@@ -116,17 +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 = "973f49cde88a582fb65755cc572ebcf6fb93ccf7" }
+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/CookiePieWw/jsonb.git", rev = "d0166c130fce903bf6c58643417a3173a6172d31", default-features = false }
+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 = "80eb97c24c88af4dd9a86f8bbaf50e741d4eb8cd" }
+meter-core = { git = "https://github.com/GreptimeTeam/greptime-meter.git", rev = "a10facb353b41460eeb98578868ebf19c2084fac" }
 mockall = "0.11.4"
 moka = "0.12"
 notify = "6.1"
@@ -137,17 +142,20 @@ opentelemetry-proto = { version = "0.5", features = [
     "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",
@@ -161,28 +169,31 @@ 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"
-similar-asserts = "1.6.0"
 # on branch v0.44.x
 sqlparser = { git = "https://github.com/GreptimeTeam/sqlparser-rs.git", rev = "54a267ac89c09b11c0c88934690530807185d3e7", features = [
     "visitor",
+    "serde",
 ] }
 strum = { version = "0.25", features = ["derive"] }
 tempfile = "3"
-tokio = { version = "1.36", features = ["full"] }
+tokio = { version = "1.40", features = ["full"] }
 tokio-postgres = "0.7"
-tokio-stream = { version = "0.1" }
+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"
 
@@ -191,6 +202,7 @@ 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" }
@@ -207,7 +219,9 @@ 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" }
@@ -224,6 +238,7 @@ file-engine = { path = "src/file-engine" }
 flow = { path = "src/flow" }
 frontend = { path = "src/frontend", default-features = false }
 index = { path = "src/index" }
+log-query = { path = "src/log-query" }
 log-store = { path = "src/log-store" }
 meta-client = { path = "src/meta-client" }
 meta-srv = { path = "src/meta-srv" }
@@ -245,9 +260,20 @@ 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 = "80eb97c24c88af4dd9a86f8bbaf50e741d4eb8cd"
+rev = "a10facb353b41460eeb98578868ebf19c2084fac"
 
 [profile.release]
 debug = 1

Makefile

@@ -8,7 +8,7 @@ CARGO_BUILD_OPTS := --locked
 IMAGE_REGISTRY ?= docker.io
 IMAGE_NAMESPACE ?= greptime
 IMAGE_TAG ?= latest
-DEV_BUILDER_IMAGE_TAG ?= 2024-06-06-b4b105ad-20240827021230
+DEV_BUILDER_IMAGE_TAG ?= 2024-10-19-a5c00e85-20241024184445
 BUILDX_MULTI_PLATFORM_BUILD ?= false
 BUILDX_BUILDER_NAME ?= gtbuilder
 BASE_IMAGE ?= ubuntu

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/flow-computation/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/protocols/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
@@ -115,7 +138,7 @@ Check the prerequisite:
 
 * [Rust toolchain](https://www.rust-lang.org/tools/install) (nightly)
 * [Protobuf compiler](https://grpc.io/docs/protoc-installation/) (>= 3.15)
-* Python toolchain (optional): Required only if built with PyO3 backend. More detail for compiling with PyO3 can be found in its [documentation](https://pyo3.rs/v0.18.1/building_and_distribution#configuring-the-python-version).
+* Python toolchain (optional): Required only if built with PyO3 backend. More details for compiling with PyO3 can be found in its [documentation](https://pyo3.rs/v0.18.1/building_and_distribution#configuring-the-python-version).
 
 Build GreptimeDB binary:
 
@@ -129,7 +152,11 @@ Run a standalone server:
 cargo run -- standalone start
 ```
 
-## Extension
+## Tools & Extensions
+
+### Kubernetes
+
+- [GreptimeDB Operator](https://github.com/GrepTimeTeam/greptimedb-operator)
 
 ### Dashboard
 
@@ -146,14 +173,19 @@ 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
 
@@ -172,12 +204,12 @@ In addition, you may:
 - Connect us with [Linkedin](https://www.linkedin.com/company/greptime/)
 - Follow us on [Twitter](https://twitter.com/greptime)
 
-## Commerial Support
+## Commercial Support
 
 If you are running GreptimeDB OSS in your organization, we offer additional
-enterprise addons, installation service, training and consulting. [Contact
+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 commerial license.
+detail of our commercial license.
 
 ## License
 

config/config.md

@@ -13,10 +13,11 @@
 | Key | Type | Default | Descriptions |
 | --- | -----| ------- | ----------- |
 | `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 | 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. |
+| `enable_telemetry` | Bool | `true` | Enable telemetry to collect anonymous usage data. Enabled by default. |
 | `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. |
@@ -60,9 +61,9 @@
 | `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 | 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`**. |
+| `wal.file_size` | String | `128MB` | The size of the WAL segment file.<br/>**It's only used when the provider is `raft_engine`**. |
+| `wal.purge_threshold` | String | `1GB` | 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 | `1m` | The interval to trigger a flush.<br/>**It's only used when the provider is `raft_engine`**. |
 | `wal.read_batch_size` | Integer | `128` | The read batch size.<br/>**It's only used when the provider is `raft_engine`**. |
 | `wal.sync_write` | Bool | `false` | Whether to use sync write.<br/>**It's only used when the provider is `raft_engine`**. |
 | `wal.enable_log_recycle` | Bool | `true` | Whether to reuse logically truncated log files.<br/>**It's only used when the provider is `raft_engine`**. |
@@ -82,6 +83,7 @@
 | `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. |
@@ -91,8 +93,8 @@
 | `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 | Unset | Cache configuration for object storage such as 'S3' etc.<br/>The local file cache directory. |
-| `storage.cache_capacity` | String | Unset | The local file cache capacity in bytes. |
+| `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`**. |
@@ -107,6 +109,11 @@
 | `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. |
@@ -114,7 +121,9 @@
 | `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 | 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`. |
@@ -122,12 +131,11 @@
 | `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. |
-| `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.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). |
@@ -142,6 +150,7 @@
 | `region_engine.mito.inverted_index.intermediate_path` | String | `""` | Deprecated, use `region_engine.mito.index.aux_path` instead. |
 | `region_engine.mito.inverted_index.metadata_cache_size` | String | `64MiB` | Cache size for inverted index metadata. |
 | `region_engine.mito.inverted_index.content_cache_size` | String | `128MiB` | Cache size for inverted index content. |
+| `region_engine.mito.inverted_index.content_cache_page_size` | String | `8MiB` | Page size for inverted index content cache. |
 | `region_engine.mito.fulltext_index` | -- | -- | The options for full-text index in Mito engine. |
 | `region_engine.mito.fulltext_index.create_on_flush` | String | `auto` | Whether to create the index on flush.<br/>- `auto`: automatically (default)<br/>- `disable`: never |
 | `region_engine.mito.fulltext_index.create_on_compaction` | String | `auto` | Whether to create the index on compaction.<br/>- `auto`: automatically (default)<br/>- `disable`: never |
@@ -160,8 +169,13 @@
 | `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. |
@@ -246,8 +260,13 @@
 | `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. |
@@ -267,13 +286,13 @@
 | `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` | Store server address default to etcd store. |
+| `store_addrs` | Array | -- | Store server address default to etcd store. |
+| `store_key_prefix` | String | `""` | If it's not empty, the metasrv will store all data with this key prefix. |
+| `backend` | String | `EtcdStore` | The datastore for meta server. |
 | `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. |
+| `enable_telemetry` | Bool | `true` | Whether to enable greptimedb telemetry. Enabled by default. |
 | `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. |
@@ -311,8 +330,13 @@
 | `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. |
@@ -333,13 +357,14 @@
 | `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. |
+| `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. |
+| `enable_telemetry` | Bool | `true` | Enable telemetry to collect anonymous usage data. Enabled by default. |
 | `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. |
@@ -374,9 +399,9 @@
 | `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 | 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`**. |
+| `wal.file_size` | String | `128MB` | The size of the WAL segment file.<br/>**It's only used when the provider is `raft_engine`**. |
+| `wal.purge_threshold` | String | `1GB` | 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 | `1m` | The interval to trigger a flush.<br/>**It's only used when the provider is `raft_engine`**. |
 | `wal.read_batch_size` | Integer | `128` | The read batch size.<br/>**It's only used when the provider is `raft_engine`**. |
 | `wal.sync_write` | Bool | `false` | Whether to use sync write.<br/>**It's only used when the provider is `raft_engine`**. |
 | `wal.enable_log_recycle` | Bool | `true` | Whether to reuse logically truncated log files.<br/>**It's only used when the provider is `raft_engine`**. |
@@ -392,11 +417,12 @@
 | `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 | Unset | Cache configuration for object storage such as 'S3' etc.<br/>The local file cache directory. |
-| `storage.cache_capacity` | String | Unset | The local file cache capacity in bytes. |
+| `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`**. |
@@ -411,6 +437,11 @@
 | `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. |
@@ -418,7 +449,9 @@
 | `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 | 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` |
@@ -426,12 +459,11 @@
 | `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. |
-| `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.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). |
@@ -444,6 +476,9 @@
 | `region_engine.mito.inverted_index.apply_on_query` | String | `auto` | Whether to apply the index on query<br/>- `auto`: automatically (default)<br/>- `disable`: never |
 | `region_engine.mito.inverted_index.mem_threshold_on_create` | String | `auto` | Memory threshold for performing an external sort during index creation.<br/>- `auto`: automatically determine the threshold based on the system memory size (default)<br/>- `unlimited`: no memory limit<br/>- `[size]` e.g. `64MB`: fixed memory threshold |
 | `region_engine.mito.inverted_index.intermediate_path` | String | `""` | Deprecated, use `region_engine.mito.index.aux_path` instead. |
+| `region_engine.mito.inverted_index.metadata_cache_size` | String | `64MiB` | Cache size for inverted index metadata. |
+| `region_engine.mito.inverted_index.content_cache_size` | String | `128MiB` | Cache size for inverted index content. |
+| `region_engine.mito.inverted_index.content_cache_page_size` | String | `8MiB` | Page size for inverted index content cache. |
 | `region_engine.mito.fulltext_index` | -- | -- | The options for full-text index in Mito engine. |
 | `region_engine.mito.fulltext_index.create_on_flush` | String | `auto` | Whether to create the index on flush.<br/>- `auto`: automatically (default)<br/>- `disable`: never |
 | `region_engine.mito.fulltext_index.create_on_compaction` | String | `auto` | Whether to create the index on compaction.<br/>- `auto`: automatically (default)<br/>- `disable`: never |
@@ -462,8 +497,13 @@
 | `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. |
@@ -508,7 +548,12 @@
 | `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 | Unset | The tokio console address. |

config/datanode.example.toml

@@ -13,12 +13,12 @@ require_lease_before_startup = false
 ## By default, it provides services after all regions have been initialized.
 init_regions_in_background = false
 
-## Enable telemetry to collect anonymous usage data.
-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
 rpc_addr = "127.0.0.1:3001"
@@ -39,6 +39,8 @@ rpc_max_recv_message_size = "512MB"
 ## @toml2docs:none-default
 rpc_max_send_message_size = "512MB"
 
+## Enable telemetry to collect anonymous usage data. Enabled by default.
+#+ enable_telemetry = true
 
 ## The HTTP server options.
 [http]
@@ -140,15 +142,15 @@ dir = "/tmp/greptimedb/wal"
 
 ## The size of the WAL segment file.
 ## **It's only used when the provider is `raft_engine`**.
-file_size = "256MB"
+file_size = "128MB"
 
 ## The threshold of the WAL size to trigger a flush.
 ## **It's only used when the provider is `raft_engine`**.
-purge_threshold = "4GB"
+purge_threshold = "1GB"
 
 ## The interval to trigger a flush.
 ## **It's only used when the provider is `raft_engine`**.
-purge_interval = "10m"
+purge_interval = "1m"
 
 ## The read batch size.
 ## **It's only used when the provider is `raft_engine`**.
@@ -210,6 +212,17 @@ create_index = true
 ## **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:
@@ -280,14 +293,14 @@ 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.
+## 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 = "/path/local_cache"
+#+ cache_path = ""
 
-## The local file cache capacity in bytes.
+## 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 = "256MB"
+cache_capacity = "5GiB"
 
 ## The S3 bucket name.
 ## **It's only used when the storage type is `S3`, `Oss` and `Gcs`**.
@@ -361,6 +374,23 @@ endpoint = "https://s3.amazonaws.com"
 ## @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"
@@ -402,8 +432,17 @@ 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"
@@ -436,14 +475,14 @@ auto_flush_interval = "1h"
 ## @toml2docs:none-default="Auto"
 #+ selector_result_cache_size = "512MB"
 
-## Whether to enable the experimental write cache.
+## 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.
 ## @toml2docs:none-default
@@ -452,12 +491,6 @@ 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
 
@@ -510,6 +543,15 @@ mem_threshold_on_create = "auto"
 ## Deprecated, use `region_engine.mito.index.aux_path` instead.
 intermediate_path = ""
 
+## Cache size for inverted index metadata.
+metadata_cache_size = "64MiB"
+
+## Cache size for inverted index content.
+content_cache_size = "128MiB"
+
+## Page size for inverted index content cache.
+content_cache_page_size = "8MiB"
+
 ## The options for full-text index in Mito engine.
 [region_engine.mito.fulltext_index]
 
@@ -577,12 +619,28 @@ 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]
@@ -607,7 +665,7 @@ url = ""
 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"
+#+ tokio_console_addr = "127.0.0.1"

config/flownode.example.toml

@@ -78,15 +78,31 @@ 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 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"
+#+ tokio_console_addr = "127.0.0.1"
 

config/frontend.example.toml

@@ -185,12 +185,28 @@ 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]
@@ -215,7 +231,7 @@ url = ""
 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"
+#+ tokio_console_addr = "127.0.0.1"

config/metasrv.example.toml

@@ -8,7 +8,13 @@ bind_addr = "127.0.0.1:3002"
 server_addr = "127.0.0.1:3002"
 
 ## Store server address default to etcd store.
-store_addr = "127.0.0.1:2379"
+store_addrs = ["127.0.0.1:2379"]
+
+## If it's not empty, the metasrv will store all data with this key prefix.
+store_key_prefix = ""
+
+## The datastore for meta server.
+backend = "EtcdStore"
 
 ## Datanode selector type.
 ## - `round_robin` (default value)
@@ -20,20 +26,14 @@ selector = "round_robin"
 ## Store data in memory.
 use_memory_store = false
 
-## Whether to enable greptimedb telemetry.
-enable_telemetry = true
-
-## If it's not empty, the metasrv will store all data with this key prefix.
-store_key_prefix = ""
-
 ## Whether to enable region failover.
 ## This feature is only available on GreptimeDB running on cluster mode and
 ## - Using Remote WAL
 ## - Using shared storage (e.g., s3).
 enable_region_failover = false
 
-## The datastore for meta server.
-backend = "EtcdStore"
+## Whether to enable greptimedb telemetry. Enabled by default.
+#+ enable_telemetry = true
 
 ## The runtime options.
 #+ [runtime]
@@ -172,12 +172,28 @@ 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]
@@ -202,7 +218,7 @@ url = ""
 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"
+#+ tokio_console_addr = "127.0.0.1"

config/standalone.example.toml

@@ -1,9 +1,6 @@
 ## The running mode of the datanode. It can be `standalone` or `distributed`.
 mode = "standalone"
 
-## Enable telemetry to collect anonymous usage data.
-enable_telemetry = true
-
 ## The default timezone of the server.
 ## @toml2docs:none-default
 default_timezone = "UTC"
@@ -15,6 +12,12 @@ 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
+
+## Enable telemetry to collect anonymous usage data. Enabled by default.
+#+ enable_telemetry = true
+
 ## The runtime options.
 #+ [runtime]
 ## The number of threads to execute the runtime for global read operations.
@@ -144,15 +147,15 @@ dir = "/tmp/greptimedb/wal"
 
 ## The size of the WAL segment file.
 ## **It's only used when the provider is `raft_engine`**.
-file_size = "256MB"
+file_size = "128MB"
 
 ## The threshold of the WAL size to trigger a flush.
 ## **It's only used when the provider is `raft_engine`**.
-purge_threshold = "4GB"
+purge_threshold = "1GB"
 
 ## The interval to trigger a flush.
 ## **It's only used when the provider is `raft_engine`**.
-purge_interval = "10m"
+purge_interval = "1m"
 
 ## The read batch size.
 ## **It's only used when the provider is `raft_engine`**.
@@ -234,6 +237,17 @@ 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:
@@ -318,14 +332,14 @@ 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.
+## 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 = "/path/local_cache"
+#+ cache_path = ""
 
-## The local file cache capacity in bytes.
+## 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 = "256MB"
+cache_capacity = "5GiB"
 
 ## The S3 bucket name.
 ## **It's only used when the storage type is `S3`, `Oss` and `Gcs`**.
@@ -399,6 +413,23 @@ endpoint = "https://s3.amazonaws.com"
 ## @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"
@@ -440,8 +471,17 @@ 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"
@@ -474,14 +514,14 @@ auto_flush_interval = "1h"
 ## @toml2docs:none-default="Auto"
 #+ selector_result_cache_size = "512MB"
 
-## Whether to enable the experimental write cache.
+## 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.
 ## @toml2docs:none-default
@@ -490,12 +530,6 @@ 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
 
@@ -554,6 +588,9 @@ metadata_cache_size = "64MiB"
 ## Cache size for inverted index content.
 content_cache_size = "128MiB"
 
+## Page size for inverted index content cache.
+content_cache_page_size = "8MiB"
+
 ## The options for full-text index in Mito engine.
 [region_engine.mito.fulltext_index]
 
@@ -621,12 +658,28 @@ 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]
@@ -651,7 +704,7 @@ url = ""
 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"
+#+ 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

@@ -15,8 +15,8 @@ RUN apt-get update && \
 RUN apt-get update && DEBIAN_FRONTEND=noninteractive apt-get install -y \
     libssl-dev \
     tzdata \
-    protobuf-compiler \
     curl \
+    unzip \
     ca-certificates \
     git \
     build-essential \
@@ -24,6 +24,29 @@ RUN apt-get update && DEBIAN_FRONTEND=noninteractive apt-get install -y \
     python3.10 \
     python3.10-dev
 
+ARG TARGETPLATFORM
+RUN echo "target platform: $TARGETPLATFORM"
+
+# Install protobuf, because the one in the apt is too old (v3.12).
+RUN if [ "$TARGETPLATFORM" = "linux/arm64" ]; then \
+    curl -OL https://github.com/protocolbuffers/protobuf/releases/download/v29.1/protoc-29.1-linux-aarch_64.zip && \
+    unzip protoc-29.1-linux-aarch_64.zip -d protoc3; \
+elif [ "$TARGETPLATFORM" = "linux/amd64" ]; then \
+    curl -OL https://github.com/protocolbuffers/protobuf/releases/download/v29.1/protoc-29.1-linux-x86_64.zip && \
+    unzip protoc-29.1-linux-x86_64.zip -d protoc3; \
+fi
+RUN mv protoc3/bin/* /usr/local/bin/
+RUN mv protoc3/include/* /usr/local/include/
+
+# 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 && \
@@ -40,7 +63,7 @@ RUN apt-get -y purge python3.8 && \
 # wildcard here. However, that requires the git's config files and the submodules all owned by the very same user.
 # It's troublesome to do this since the dev build runs in Docker, which is under user "root"; while outside the Docker,
 # it can be a different user that have prepared the submodules.
-RUN git config --global --add safe.directory *
+RUN git config --global --add safe.directory '*'
 
 # Install Python dependencies.
 COPY $DOCKER_BUILD_ROOT/docker/python/requirements.txt /etc/greptime/requirements.txt
@@ -57,7 +80,9 @@ 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

@@ -48,4 +48,4 @@ Please refer to [SQL query](./query.sql) for GreptimeDB and Clickhouse, and [que
 
 ## 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/operations/configuration/#storage-options).
+- You can setup GreptimeDB to use S3 as storage, see [here](https://docs.greptime.com/user-guide/deployments/configuration#storage-options).

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`

rust-toolchain.toml

@@ -1,3 +1,3 @@
 [toolchain]
-channel = "nightly-2024-06-06"
-
+channel = "nightly-2024-10-19"
+components = ["rust-analyzer"]

scripts/check-snafu.py

@@ -58,8 +58,10 @@ def main():
         if not check_snafu_in_files(branch_name, other_rust_files)
     ]
 
-    for name in unused_snafu:
-        print(name)
+    if unused_snafu:
+        print("Unused error variants:")
+        for name in unused_snafu:
+            print(name)
 
     if unused_snafu:
         raise SystemExit(1)

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.

shell.nix

@@ -0,0 +1,27 @@
+let
+  nixpkgs = fetchTarball "https://github.com/NixOS/nixpkgs/tarball/nixos-unstable";
+  fenix = import (fetchTarball "https://github.com/nix-community/fenix/archive/main.tar.gz") {};
+  pkgs = import nixpkgs { config = {}; overlays = []; };
+in
+
+pkgs.mkShell rec {
+  nativeBuildInputs = with pkgs; [
+    pkg-config
+    git
+    clang
+    gcc
+    protobuf
+    mold
+    (fenix.fromToolchainFile {
+      dir = ./.;
+    })
+    cargo-nextest
+    taplo
+  ];
+
+  buildInputs = with pkgs; [
+    libgit2
+  ];
+
+  LD_LIBRARY_PATH = pkgs.lib.makeLibraryPath buildInputs;
+}

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::{
@@ -35,15 +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, JsonTypeExtension, QueryRequest,
-    Row, SemanticType,
+    self, ColumnDataTypeExtension, DdlRequest, DecimalTypeExtension, JsonTypeExtension,
+    QueryRequest, Row, SemanticType, VectorTypeExtension,
 };
 use paste::paste;
 use snafu::prelude::*;
@@ -115,6 +115,7 @@ impl From<ColumnDataTypeWrapper> for ConcreteDataType {
                     ConcreteDataType::binary_datatype()
                 }
             }
+            ColumnDataType::Json => ConcreteDataType::json_datatype(),
             ColumnDataType::String => ConcreteDataType::string_datatype(),
             ColumnDataType::Date => ConcreteDataType::date_datatype(),
             ColumnDataType::Datetime => ConcreteDataType::datetime_datatype(),
@@ -148,6 +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()
+                }
+            }
         }
     }
 }
@@ -229,6 +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 {
@@ -247,7 +268,7 @@ impl TryFrom<ConcreteDataType> for ColumnDataTypeWrapper {
             ConcreteDataType::UInt64(_) => ColumnDataType::Uint64,
             ConcreteDataType::Float32(_) => ColumnDataType::Float32,
             ConcreteDataType::Float64(_) => ColumnDataType::Float64,
-            ConcreteDataType::Binary(_) | ConcreteDataType::Json(_) => ColumnDataType::Binary,
+            ConcreteDataType::Binary(_) => ColumnDataType::Binary,
             ConcreteDataType::String(_) => ColumnDataType::String,
             ConcreteDataType::Date(_) => ColumnDataType::Date,
             ConcreteDataType::DateTime(_) => ColumnDataType::Datetime,
@@ -269,6 +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(_)
@@ -287,15 +310,17 @@ impl TryFrom<ConcreteDataType> for ColumnDataTypeWrapper {
                         })),
                     })
             }
-            ColumnDataType::Binary => {
-                if datatype == ConcreteDataType::json_datatype() {
-                    // Json is the same as  binary in proto. The extension marks the binary in proto is actually a json.
-                    Some(ColumnDataTypeExtension {
-                        type_ext: Some(TypeExt::JsonType(JsonTypeExtension::JsonBinary.into())),
+            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 _,
+                        })),
                     })
-                } else {
-                    None
-                }
             }
             _ => None,
         };
@@ -416,6 +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()
+        },
     }
 }
 
@@ -456,13 +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!(),
     });
@@ -496,25 +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,
     }
 }
 
@@ -562,11 +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
@@ -657,7 +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()
                     }),
                 ))
             }
@@ -667,6 +701,7 @@ 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(_)
@@ -802,18 +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,
@@ -832,6 +867,7 @@ 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(_)
@@ -856,10 +892,7 @@ pub fn is_column_type_value_eq(
     ColumnDataTypeWrapper::try_new(type_value, type_extension)
         .map(|wrapper| {
             let datatype = ConcreteDataType::from(wrapper);
-            (datatype == *expect_type)
-            // Json type leverage binary type in pb, so this is valid.
-                || (datatype == ConcreteDataType::binary_datatype()
-                    && *expect_type == ConcreteDataType::json_datatype())
+            expect_type == &datatype
         })
         .unwrap_or(false)
 }
@@ -941,18 +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))),
@@ -1044,13 +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!(),
         },
@@ -1061,6 +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,
@@ -1149,6 +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]
@@ -1236,7 +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]
@@ -1332,6 +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());
@@ -1506,11 +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]
@@ -1590,9 +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);
 
@@ -1605,9 +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);
 
@@ -1636,9 +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/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, SKIPPING_INDEX_KEY,
 };
+use greptime_proto::v1::Analyzer;
 use snafu::ResultExt;
 
 use crate::error::{self, Result};
@@ -25,6 +27,10 @@ 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";
+/// Key used to store skip index options in gRPC column options.
+const SKIPPING_INDEX_GRPC_KEY: &str = "skipping_index";
 
 /// Tries to construct a `ColumnSchema` from the given  `ColumnDef`.
 pub fn try_as_column_schema(column_def: &ColumnDef) -> Result<ColumnSchema> {
@@ -49,10 +55,16 @@ 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());
+        }
+        if let Some(skipping_index) = options.options.get(SKIPPING_INDEX_GRPC_KEY) {
+            metadata.insert(SKIPPING_INDEX_KEY.to_string(), skipping_index.clone());
+        }
     }
 
     ColumnSchema::new(&column_def.name, data_type.into(), column_def.is_nullable)
@@ -70,7 +82,17 @@ 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());
+    }
+    if let Some(skipping_index) = column_schema.metadata().get(SKIPPING_INDEX_KEY) {
+        options
+            .options
+            .insert(SKIPPING_INDEX_GRPC_KEY.to_string(), skipping_index.clone());
     }
 
     (!options.options.is_empty()).then_some(options)
@@ -93,6 +115,14 @@ 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 +145,13 @@ 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 +172,7 @@ mod tests {
                 ..Default::default()
             }
         );
+        assert!(schema.is_inverted_indexed());
     }
 
     #[test]
@@ -153,12 +187,17 @@ 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/permission.rs

@@ -25,6 +25,7 @@ pub enum PermissionReq<'a> {
     GrpcRequest(&'a Request),
     SqlStatement(&'a Statement),
     PromQuery,
+    LogQuery,
     Opentsdb,
     LineProtocol,
     PromStoreWrite,

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/cache/Cargo.toml

@@ -11,4 +11,3 @@ common-macro.workspace = true
 common-meta.workspace = true
 moka.workspace = true
 snafu.workspace = true
-substrait.workspace = true

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

@@ -18,12 +18,13 @@ async-stream.workspace = true
 async-trait = "0.1"
 bytes.workspace = true
 common-catalog.workspace = true
-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 +49,7 @@ sql.workspace = true
 store-api.workspace = true
 table.workspace = true
 tokio.workspace = true
+tokio-stream = "0.1"
 
 [dev-dependencies]
 cache.workspace = true
@@ -55,7 +57,5 @@ catalog = { workspace = true, features = ["testing"] }
 chrono.workspace = true
 common-meta = { workspace = true, features = ["testing"] }
 common-query = { workspace = true, features = ["testing"] }
-common-test-util.workspace = true
-log-store.workspace = true
 object-store.workspace = true
 tokio.workspace = true

src/catalog/src/error.rs

@@ -50,13 +50,27 @@ 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 flow stats"))]
+    ListFlowStats {
+        #[snafu(implicit)]
+        location: Location,
+        source: BoxedError,
+    },
+
     #[snafu(display("Failed to list flows in catalog {catalog}"))]
     ListFlows {
         #[snafu(implicit)]
@@ -82,6 +96,32 @@ 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)]
@@ -145,6 +185,12 @@ 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 },
 
@@ -266,7 +312,10 @@ 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,
 
@@ -283,7 +332,11 @@ impl ErrorExt for Error {
             | Error::ListNodes { source, .. }
             | Error::ListSchemas { source, .. }
             | Error::ListTables { source, .. }
-            | Error::ListFlows { source, .. } => source.status_code(),
+            | Error::ListFlows { source, .. }
+            | Error::ListFlowStats { source, .. }
+            | Error::ListProcedures { source, .. }
+            | Error::ListRegionStats { source, .. }
+            | Error::ConvertProtoData { source, .. } => source.status_code(),
 
             Error::CreateTable { source, .. } => source.status_code(),
 

src/catalog/src/information_extension.rs

@@ -0,0 +1,101 @@
+// 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::key::flow::flow_state::FlowStat;
+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)
+    }
+
+    async fn flow_stats(&self) -> std::result::Result<Option<FlowStat>, Self::Error> {
+        self.meta_client
+            .list_flow_stats()
+            .await
+            .map_err(BoxedError::new)
+            .context(crate::error::ListFlowStatsSnafu)
+    }
+}

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

@@ -22,6 +22,7 @@ use common_error::ext::BoxedError;
 use common_meta::cache_invalidator::KvCacheInvalidator;
 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
     }
@@ -305,7 +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;
@@ -313,7 +321,7 @@ impl KvCacheInvalidator for CachedMetaKvBackend {
     }
 }
 
-impl CachedMetaKvBackend {
+impl CachedKvBackend {
     // only for test
     #[cfg(test)]
     fn wrap(kv_backend: KvBackendRef) -> Self {
@@ -466,7 +474,7 @@ mod tests {
     use common_meta::rpc::KeyValue;
     use dashmap::DashMap;
 
-    use super::CachedMetaKvBackend;
+    use super::CachedKvBackend;
 
     #[derive(Default)]
     pub struct SimpleKvBackend {
@@ -540,7 +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;
@@ -31,9 +30,9 @@ 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};
@@ -42,12 +41,14 @@ 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;
@@ -60,27 +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
@@ -104,23 +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 {
@@ -130,6 +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]
@@ -179,21 +184,18 @@ impl CatalogManager for KvBackendCatalogManager {
         schema: &str,
         query_ctx: Option<&QueryContext>,
     ) -> Result<Vec<String>> {
-        let stream = self
+        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, query_ctx));
+            .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> {
@@ -303,36 +305,68 @@ impl CatalogManager for KvBackendCatalogManager {
             }
         });
 
-        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))
     }
 }

src/catalog/src/kvbackend/table_cache.rs

@@ -38,7 +38,7 @@ pub fn new_table_cache(
 ) -> TableCache {
     let init = init_factory(table_info_cache, table_name_cache);
 
-    CacheContainer::new(name, cache, Box::new(invalidator), init, Box::new(filter))
+    CacheContainer::new(name, cache, Box::new(invalidator), init, filter)
 }
 
 fn init_factory(

src/catalog/src/lib.rs

@@ -30,6 +30,7 @@ use table::TableRef;
 use crate::error::Result;
 
 pub mod error;
+pub mod information_extension;
 pub mod kvbackend;
 pub mod memory;
 mod metrics;

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 +32,12 @@ 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::flow_state::FlowStat;
 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 +50,7 @@ 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;
@@ -186,8 +193,19 @@ impl SystemSchemaProviderInner for InformationSchemaProvider {
             )) as _),
             FLOWS => Some(Arc::new(InformationSchemaFlows::new(
                 self.catalog_name.clone(),
+                self.catalog_manager.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 +253,14 @@ 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 +276,6 @@ 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 +324,46 @@ 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>;
+
+    /// Get the flow statistics. If no flownode is available, return `None`.
+    async fn flow_stats(&self) -> std::result::Result<Option<FlowStat>, 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![])
+    }
+
+    async fn flow_stats(&self) -> std::result::Result<Option<FlowStat>, Self::Error> {
+        Ok(None)
+    }
+}

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/flows.rs

@@ -12,11 +12,12 @@
 // See the License for the specific language governing permissions and
 // limitations under the License.
 
-use std::sync::Arc;
+use std::sync::{Arc, Weak};
 
 use common_catalog::consts::INFORMATION_SCHEMA_FLOW_TABLE_ID;
 use common_error::ext::BoxedError;
 use common_meta::key::flow::flow_info::FlowInfoValue;
+use common_meta::key::flow::flow_state::FlowStat;
 use common_meta::key::flow::FlowMetadataManager;
 use common_meta::key::FlowId;
 use common_recordbatch::adapter::RecordBatchStreamAdapter;
@@ -28,7 +29,9 @@ use datatypes::prelude::ConcreteDataType as CDT;
 use datatypes::scalars::ScalarVectorBuilder;
 use datatypes::schema::{ColumnSchema, Schema, SchemaRef};
 use datatypes::value::Value;
-use datatypes::vectors::{Int64VectorBuilder, StringVectorBuilder, UInt32VectorBuilder, VectorRef};
+use datatypes::vectors::{
+    Int64VectorBuilder, StringVectorBuilder, UInt32VectorBuilder, UInt64VectorBuilder, VectorRef,
+};
 use futures::TryStreamExt;
 use snafu::{OptionExt, ResultExt};
 use store_api::storage::{ScanRequest, TableId};
@@ -38,6 +41,8 @@ use crate::error::{
 };
 use crate::information_schema::{Predicates, FLOWS};
 use crate::system_schema::information_schema::InformationTable;
+use crate::system_schema::utils;
+use crate::CatalogManager;
 
 const INIT_CAPACITY: usize = 42;
 
@@ -45,6 +50,7 @@ const INIT_CAPACITY: usize = 42;
 // pk is (flow_name, flow_id, table_catalog)
 pub const FLOW_NAME: &str = "flow_name";
 pub const FLOW_ID: &str = "flow_id";
+pub const STATE_SIZE: &str = "state_size";
 pub const TABLE_CATALOG: &str = "table_catalog";
 pub const FLOW_DEFINITION: &str = "flow_definition";
 pub const COMMENT: &str = "comment";
@@ -55,20 +61,24 @@ pub const FLOWNODE_IDS: &str = "flownode_ids";
 pub const OPTIONS: &str = "options";
 
 /// The `information_schema.flows` to provides information about flows in databases.
+///
 pub(super) struct InformationSchemaFlows {
     schema: SchemaRef,
     catalog_name: String,
+    catalog_manager: Weak<dyn CatalogManager>,
     flow_metadata_manager: Arc<FlowMetadataManager>,
 }
 
 impl InformationSchemaFlows {
     pub(super) fn new(
         catalog_name: String,
+        catalog_manager: Weak<dyn CatalogManager>,
         flow_metadata_manager: Arc<FlowMetadataManager>,
     ) -> Self {
         Self {
             schema: Self::schema(),
             catalog_name,
+            catalog_manager,
             flow_metadata_manager,
         }
     }
@@ -80,6 +90,7 @@ impl InformationSchemaFlows {
             vec![
                 (FLOW_NAME, CDT::string_datatype(), false),
                 (FLOW_ID, CDT::uint32_datatype(), false),
+                (STATE_SIZE, CDT::uint64_datatype(), true),
                 (TABLE_CATALOG, CDT::string_datatype(), false),
                 (FLOW_DEFINITION, CDT::string_datatype(), false),
                 (COMMENT, CDT::string_datatype(), true),
@@ -99,6 +110,7 @@ impl InformationSchemaFlows {
         InformationSchemaFlowsBuilder::new(
             self.schema.clone(),
             self.catalog_name.clone(),
+            self.catalog_manager.clone(),
             &self.flow_metadata_manager,
         )
     }
@@ -144,10 +156,12 @@ impl InformationTable for InformationSchemaFlows {
 struct InformationSchemaFlowsBuilder {
     schema: SchemaRef,
     catalog_name: String,
+    catalog_manager: Weak<dyn CatalogManager>,
     flow_metadata_manager: Arc<FlowMetadataManager>,
 
     flow_names: StringVectorBuilder,
     flow_ids: UInt32VectorBuilder,
+    state_sizes: UInt64VectorBuilder,
     table_catalogs: StringVectorBuilder,
     raw_sqls: StringVectorBuilder,
     comments: StringVectorBuilder,
@@ -162,15 +176,18 @@ impl InformationSchemaFlowsBuilder {
     fn new(
         schema: SchemaRef,
         catalog_name: String,
+        catalog_manager: Weak<dyn CatalogManager>,
         flow_metadata_manager: &Arc<FlowMetadataManager>,
     ) -> Self {
         Self {
             schema,
             catalog_name,
+            catalog_manager,
             flow_metadata_manager: flow_metadata_manager.clone(),
 
             flow_names: StringVectorBuilder::with_capacity(INIT_CAPACITY),
             flow_ids: UInt32VectorBuilder::with_capacity(INIT_CAPACITY),
+            state_sizes: UInt64VectorBuilder::with_capacity(INIT_CAPACITY),
             table_catalogs: StringVectorBuilder::with_capacity(INIT_CAPACITY),
             raw_sqls: StringVectorBuilder::with_capacity(INIT_CAPACITY),
             comments: StringVectorBuilder::with_capacity(INIT_CAPACITY),
@@ -195,6 +212,11 @@ impl InformationSchemaFlowsBuilder {
             .flow_names(&catalog_name)
             .await;
 
+        let flow_stat = {
+            let information_extension = utils::information_extension(&self.catalog_manager)?;
+            information_extension.flow_stats().await?
+        };
+
         while let Some((flow_name, flow_id)) = stream
             .try_next()
             .await
@@ -213,7 +235,7 @@ impl InformationSchemaFlowsBuilder {
                     catalog_name: catalog_name.to_string(),
                     flow_name: flow_name.to_string(),
                 })?;
-            self.add_flow(&predicates, flow_id.flow_id(), flow_info)?;
+            self.add_flow(&predicates, flow_id.flow_id(), flow_info, &flow_stat)?;
         }
 
         self.finish()
@@ -224,6 +246,7 @@ impl InformationSchemaFlowsBuilder {
         predicates: &Predicates,
         flow_id: FlowId,
         flow_info: FlowInfoValue,
+        flow_stat: &Option<FlowStat>,
     ) -> Result<()> {
         let row = [
             (FLOW_NAME, &Value::from(flow_info.flow_name().to_string())),
@@ -238,6 +261,11 @@ impl InformationSchemaFlowsBuilder {
         }
         self.flow_names.push(Some(flow_info.flow_name()));
         self.flow_ids.push(Some(flow_id));
+        self.state_sizes.push(
+            flow_stat
+                .as_ref()
+                .and_then(|state| state.state_size.get(&flow_id).map(|v| *v as u64)),
+        );
         self.table_catalogs.push(Some(flow_info.catalog_name()));
         self.raw_sqls.push(Some(flow_info.raw_sql()));
         self.comments.push(Some(flow_info.comment()));
@@ -270,6 +298,7 @@ impl InformationSchemaFlowsBuilder {
         let columns: Vec<VectorRef> = vec![
             Arc::new(self.flow_names.finish()),
             Arc::new(self.flow_ids.finish()),
+            Arc::new(self.state_sizes.finish()),
             Arc::new(self.table_catalogs.finish()),
             Arc::new(self.raw_sqls.finish()),
             Arc::new(self.comments.finish()),

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

@@ -54,6 +54,10 @@ const INIT_CAPACITY: usize = 42;
 pub(crate) const PRI_CONSTRAINT_NAME: &str = "PRIMARY";
 /// Time index constraint name
 pub(crate) const TIME_INDEX_CONSTRAINT_NAME: &str = "TIME INDEX";
+/// Inverted index constraint name
+pub(crate) const INVERTED_INDEX_CONSTRAINT_NAME: &str = "INVERTED INDEX";
+/// Fulltext index constraint name
+pub(crate) const FULLTEXT_INDEX_CONSTRAINT_NAME: &str = "FULLTEXT INDEX";
 
 /// The virtual table implementation for `information_schema.KEY_COLUMN_USAGE`.
 pub(super) struct InformationSchemaKeyColumnUsage {
@@ -216,14 +220,13 @@ impl InformationSchemaKeyColumnUsageBuilder {
             let mut stream = catalog_manager.tables(&catalog_name, &schema_name, None);
 
             while let Some(table) = stream.try_next().await? {
-                let mut primary_constraints = vec![];
-
                 let table_info = table.table_info();
                 let table_name = &table_info.name;
                 let keys = &table_info.meta.primary_key_indices;
                 let schema = table.schema();
 
                 for (idx, column) in schema.column_schemas().iter().enumerate() {
+                    let mut constraints = vec![];
                     if column.is_time_index() {
                         self.add_key_column_usage(
                             &predicates,
@@ -236,30 +239,31 @@ impl InformationSchemaKeyColumnUsageBuilder {
                             1, //always 1 for time index
                         );
                     }
-                    if keys.contains(&idx) {
-                        primary_constraints.push((
-                            catalog_name.clone(),
-                            schema_name.clone(),
-                            table_name.to_string(),
-                            column.name.clone(),
-                        ));
-                    }
                     // TODO(dimbtp): foreign key constraint not supported yet
-                }
+                    if keys.contains(&idx) {
+                        constraints.push(PRI_CONSTRAINT_NAME);
+                    }
+                    if column.is_inverted_indexed() {
+                        constraints.push(INVERTED_INDEX_CONSTRAINT_NAME);
+                    }
 
-                for (i, (catalog_name, schema_name, table_name, column_name)) in
-                    primary_constraints.into_iter().enumerate()
-                {
-                    self.add_key_column_usage(
-                        &predicates,
-                        &schema_name,
-                        PRI_CONSTRAINT_NAME,
-                        &catalog_name,
-                        &schema_name,
-                        &table_name,
-                        &column_name,
-                        i as u32 + 1,
-                    );
+                    if column.has_fulltext_index_key() {
+                        constraints.push(FULLTEXT_INDEX_CONSTRAINT_NAME);
+                    }
+
+                    if !constraints.is_empty() {
+                        let aggregated_constraints = constraints.join(", ");
+                        self.add_key_column_usage(
+                            &predicates,
+                            &schema_name,
+                            &aggregated_constraints,
+                            &catalog_name,
+                            &schema_name,
+                            table_name,
+                            &column.name,
+                            idx as u32 + 1,
+                        );
+                    }
                 }
             }
         }

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};
@@ -236,7 +235,8 @@ 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);
 
@@ -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

@@ -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

@@ -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_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);
 
+        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/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/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};
@@ -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;
 

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

@@ -259,7 +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;
@@ -269,6 +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> {
@@ -323,10 +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![]);

src/cli/Cargo.toml

@@ -0,0 +1,63 @@
+[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-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
+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

@@ -12,24 +12,35 @@
 // 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::greptime_result_v1::GreptimedbV1Response;
+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};
 
-pub(crate) struct DatabaseClient {
+#[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>) -> Self {
+    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))
@@ -41,6 +52,7 @@ impl DatabaseClient {
             addr,
             catalog,
             auth_header,
+            timeout,
         }
     }
 
@@ -63,6 +75,11 @@ impl DatabaseClient {
             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),
         })?;

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

@@ -15,9 +15,11 @@
 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};
@@ -25,11 +27,10 @@ use tokio::fs::File;
 use tokio::io::{AsyncWriteExt, BufWriter};
 use tokio::sync::Semaphore;
 use tokio::time::Instant;
-use tracing_appender::non_blocking::WorkerGuard;
 
-use crate::cli::database::DatabaseClient;
-use crate::cli::{database, Instance, Tool};
+use crate::database::DatabaseClient;
 use crate::error::{EmptyResultSnafu, Error, FileIoSnafu, Result, SchemaNotFoundSnafu};
+use crate::{database, Tool};
 
 type TableReference = (String, String, String);
 
@@ -83,28 +84,38 @@ pub struct ExportCommand {
     /// 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, guard: Vec<WorkerGuard>) -> Result<Instance> {
-        let (catalog, schema) = database::split_database(&self.database)?;
+    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());
+        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(Instance::new(
-            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(),
-            }),
-            guard,
-        ))
+        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(),
+        }))
     }
 }
 
@@ -452,97 +463,22 @@ impl Export {
 
 #[async_trait]
 impl Tool for Export {
-    async fn do_work(&self) -> Result<()> {
+    async fn do_work(&self) -> std::result::Result<(), BoxedError> {
         match self.target {
             ExportTarget::Schema => {
-                self.export_create_database().await?;
-                self.export_create_table().await
+                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,
+            ExportTarget::Data => self.export_database_data().await.map_err(BoxedError::new),
             ExportTarget::All => {
-                self.export_create_database().await?;
-                self.export_create_table().await?;
-                self.export_database_data().await
+                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)
             }
         }
     }
 }
-
-#[cfg(test)]
-mod tests {
-    use clap::Parser;
-    use client::{Client, Database};
-    use common_catalog::consts::{DEFAULT_CATALOG_NAME, DEFAULT_SCHEMA_NAME};
-    use common_telemetry::logging::LoggingOptions;
-
-    use crate::error::Result as CmdResult;
-    use crate::options::GlobalOptions;
-    use crate::{cli, standalone, App};
-
-    #[tokio::test(flavor = "multi_thread")]
-    async fn test_export_create_table_with_quoted_names() -> CmdResult<()> {
-        let output_dir = tempfile::tempdir().unwrap();
-
-        let standalone = standalone::Command::parse_from([
-            "standalone",
-            "start",
-            "--data-home",
-            &*output_dir.path().to_string_lossy(),
-        ]);
-
-        let standalone_opts = standalone.load_options(&GlobalOptions::default()).unwrap();
-        let mut instance = standalone.build(standalone_opts).await?;
-        instance.start().await?;
-
-        let client = Client::with_urls(["127.0.0.1:4001"]);
-        let database = Database::new(DEFAULT_CATALOG_NAME, DEFAULT_SCHEMA_NAME, client);
-        database
-            .sql(r#"CREATE DATABASE "cli.export.create_table";"#)
-            .await
-            .unwrap();
-        database
-            .sql(
-                r#"CREATE TABLE "cli.export.create_table"."a.b.c"(
-                        ts TIMESTAMP,
-                        TIME INDEX (ts)
-                    ) engine=mito;
-                "#,
-            )
-            .await
-            .unwrap();
-
-        let output_dir = tempfile::tempdir().unwrap();
-        let cli = cli::Command::parse_from([
-            "cli",
-            "export",
-            "--addr",
-            "127.0.0.1:4000",
-            "--output-dir",
-            &*output_dir.path().to_string_lossy(),
-            "--target",
-            "schema",
-        ]);
-        let mut cli_app = cli.build(LoggingOptions::default()).await?;
-        cli_app.start().await?;
-
-        instance.stop().await?;
-
-        let output_file = output_dir
-            .path()
-            .join("greptime")
-            .join("cli.export.create_table")
-            .join("create_tables.sql");
-        let res = std::fs::read_to_string(output_file).unwrap();
-        let expect = r#"CREATE TABLE IF NOT EXISTS "a.b.c" (
-  "ts" TIMESTAMP(3) NOT NULL,
-  TIME INDEX ("ts")
-)
-
-ENGINE=mito
-;
-"#;
-        assert_eq!(res.trim(), expect.trim());
-
-        Ok(())
-    }
-}

src/cli/src/helper.rs

@@ -19,7 +19,7 @@ use rustyline::highlight::{Highlighter, MatchingBracketHighlighter};
 use rustyline::hint::{Hinter, HistoryHinter};
 use rustyline::validate::{ValidationContext, ValidationResult, Validator};
 
-use crate::cli::cmd::ReplCommand;
+use crate::cmd::ReplCommand;
 
 pub(crate) struct RustylineHelper {
     hinter: HistoryHinter,

src/cli/src/import.rs

@@ -14,19 +14,20 @@
 
 use std::path::PathBuf;
 use std::sync::Arc;
+use std::time::Duration;
 
 use async_trait::async_trait;
 use clap::{Parser, ValueEnum};
 use common_catalog::consts::DEFAULT_SCHEMA_NAME;
+use common_error::ext::BoxedError;
 use common_telemetry::{error, info, warn};
 use snafu::{OptionExt, ResultExt};
 use tokio::sync::Semaphore;
 use tokio::time::Instant;
-use tracing_appender::non_blocking::WorkerGuard;
 
-use crate::cli::database::DatabaseClient;
-use crate::cli::{database, Instance, Tool};
+use crate::database::DatabaseClient;
 use crate::error::{Error, FileIoSnafu, Result, SchemaNotFoundSnafu};
+use crate::{database, Tool};
 
 #[derive(Debug, Default, Clone, ValueEnum)]
 enum ImportTarget {
@@ -68,25 +69,35 @@ pub struct ImportCommand {
     /// 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 ImportCommand {
-    pub async fn build(&self, guard: Vec<WorkerGuard>) -> Result<Instance> {
-        let (catalog, schema) = database::split_database(&self.database)?;
-        let database_client =
-            DatabaseClient::new(self.addr.clone(), catalog.clone(), self.auth_basic.clone());
+    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(Instance::new(
-            Box::new(Import {
-                catalog,
-                schema,
-                database_client,
-                input_dir: self.input_dir.clone(),
-                parallelism: self.import_jobs,
-                target: self.target.clone(),
-            }),
-            guard,
-        ))
+        Ok(Box::new(Import {
+            catalog,
+            schema,
+            database_client,
+            input_dir: self.input_dir.clone(),
+            parallelism: self.import_jobs,
+            target: self.target.clone(),
+        }))
     }
 }
 
@@ -205,13 +216,13 @@ impl Import {
 
 #[async_trait]
 impl Tool for Import {
-    async fn do_work(&self) -> Result<()> {
+    async fn do_work(&self) -> std::result::Result<(), BoxedError> {
         match self.target {
-            ImportTarget::Schema => self.import_create_table().await,
-            ImportTarget::Data => self.import_database_data().await,
+            ImportTarget::Schema => self.import_create_table().await.map_err(BoxedError::new),
+            ImportTarget::Data => self.import_database_data().await.map_err(BoxedError::new),
             ImportTarget::All => {
-                self.import_create_table().await?;
-                self.import_database_data().await
+                self.import_create_table().await.map_err(BoxedError::new)?;
+                self.import_database_data().await.map_err(BoxedError::new)
             }
         }
     }

src/cli/src/lib.rs

@@ -0,0 +1,60 @@
+// 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.
+
+mod bench;
+pub mod error;
+// Wait for https://github.com/GreptimeTeam/greptimedb/issues/2373
+#[allow(unused)]
+mod cmd;
+mod export;
+mod helper;
+
+// Wait for https://github.com/GreptimeTeam/greptimedb/issues/2373
+mod database;
+mod import;
+#[allow(unused)]
+mod repl;
+
+use async_trait::async_trait;
+use clap::Parser;
+use common_error::ext::BoxedError;
+pub use database::DatabaseClient;
+use error::Result;
+pub use repl::Repl;
+
+pub use crate::bench::BenchTableMetadataCommand;
+pub use crate::export::ExportCommand;
+pub use crate::import::ImportCommand;
+
+#[async_trait]
+pub trait Tool: Send + Sync {
+    async fn do_work(&self) -> std::result::Result<(), BoxedError>;
+}
+
+#[derive(Debug, Parser)]
+pub(crate) struct AttachCommand {
+    #[clap(long)]
+    pub(crate) grpc_addr: String,
+    #[clap(long)]
+    pub(crate) meta_addr: Option<String>,
+    #[clap(long, action)]
+    pub(crate) disable_helper: bool,
+}
+
+impl AttachCommand {
+    #[allow(dead_code)]
+    async fn build(self) -> Result<Box<dyn Tool>> {
+        unimplemented!("Wait for https://github.com/GreptimeTeam/greptimedb/issues/2373")
+    }
+}

src/cli/src/repl.rs

@@ -20,22 +20,23 @@ use cache::{
     build_fundamental_cache_registry, with_default_composite_cache_registry, TABLE_CACHE_NAME,
     TABLE_ROUTE_CACHE_NAME,
 };
+use catalog::information_extension::DistributedInformationExtension;
 use catalog::kvbackend::{
-    CachedMetaKvBackend, CachedMetaKvBackendBuilder, KvBackendCatalogManager, MetaKvBackend,
+    CachedKvBackend, CachedKvBackendBuilder, KvBackendCatalogManager, MetaKvBackend,
 };
 use client::{Client, Database, OutputData, DEFAULT_CATALOG_NAME, DEFAULT_SCHEMA_NAME};
 use common_base::Plugins;
 use common_config::Mode;
 use common_error::ext::ErrorExt;
 use common_meta::cache::{CacheRegistryBuilder, LayeredCacheRegistryBuilder};
+use common_meta::kv_backend::KvBackendRef;
 use common_query::Output;
 use common_recordbatch::RecordBatches;
 use common_telemetry::debug;
 use either::Either;
-use meta_client::client::MetaClientBuilder;
+use meta_client::client::{ClusterKvBackend, MetaClientBuilder};
 use query::datafusion::DatafusionQueryEngine;
 use query::parser::QueryLanguageParser;
-use query::plan::LogicalPlan;
 use query::query_engine::{DefaultSerializer, QueryEngineState};
 use query::QueryEngine;
 use rustyline::error::ReadlineError;
@@ -44,15 +45,14 @@ use session::context::QueryContext;
 use snafu::{OptionExt, ResultExt};
 use substrait::{DFLogicalSubstraitConvertor, SubstraitPlan};
 
-use crate::cli::cmd::ReplCommand;
-use crate::cli::helper::RustylineHelper;
-use crate::cli::AttachCommand;
-use crate::error;
+use crate::cmd::ReplCommand;
 use crate::error::{
     CollectRecordBatchesSnafu, ParseSqlSnafu, PlanStatementSnafu, PrettyPrintRecordBatchesSnafu,
     ReadlineSnafu, ReplCreationSnafu, RequestDatabaseSnafu, Result, StartMetaClientSnafu,
     SubstraitEncodeLogicalPlanSnafu,
 };
+use crate::helper::RustylineHelper;
+use crate::{error, AttachCommand};
 
 /// Captures the state of the repl, gathers commands and executes them one by one
 pub struct Repl {
@@ -175,11 +175,11 @@ impl Repl {
 
             let plan = query_engine
                 .planner()
-                .plan(stmt, query_ctx.clone())
+                .plan(&stmt, query_ctx.clone())
                 .await
                 .context(PlanStatementSnafu)?;
 
-            let LogicalPlan::DfPlan(plan) = query_engine
+            let plan = query_engine
                 .optimize(&query_engine.engine_context(query_ctx), &plan)
                 .context(PlanStatementSnafu)?;
 
@@ -259,8 +259,9 @@ async fn create_query_engine(meta_addr: &str) -> Result<DatafusionQueryEngine> {
         .context(StartMetaClientSnafu)?;
     let meta_client = Arc::new(meta_client);
 
-    let cached_meta_backend =
-        Arc::new(CachedMetaKvBackendBuilder::new(meta_client.clone()).build());
+    let cached_meta_backend = Arc::new(
+        CachedKvBackendBuilder::new(Arc::new(MetaKvBackend::new(meta_client.clone()))).build(),
+    );
     let layered_cache_builder = LayeredCacheRegistryBuilder::default().add_cache_registry(
         CacheRegistryBuilder::default()
             .add_cache(cached_meta_backend.clone())
@@ -276,11 +277,12 @@ async fn create_query_engine(meta_addr: &str) -> Result<DatafusionQueryEngine> {
         .build(),
     );
 
+    let information_extension = Arc::new(DistributedInformationExtension::new(meta_client.clone()));
     let catalog_manager = KvBackendCatalogManager::new(
-        Mode::Distributed,
-        Some(meta_client.clone()),
+        information_extension,
         cached_meta_backend.clone(),
         layered_cache_registry,
+        None,
     );
     let plugins: Plugins = Default::default();
     let state = Arc::new(QueryEngineState::new(

src/client/Cargo.toml

@@ -28,7 +28,7 @@ enum_dispatch = "0.3"
 futures-util.workspace = true
 lazy_static.workspace = true
 moka = { workspace = true, features = ["future"] }
-parking_lot = "0.12"
+parking_lot.workspace = true
 prometheus.workspace = true
 prost.workspace = true
 query.workspace = true
@@ -42,10 +42,7 @@ tonic.workspace = true
 
 [dev-dependencies]
 common-grpc-expr.workspace = true
-datanode.workspace = true
-derive-new = "0.5"
 tracing = "0.1"
-tracing-subscriber = { version = "0.3", features = ["env-filter"] }
 
 [dev-dependencies.substrait_proto]
 package = "substrait"

src/client/src/database.rs

@@ -18,7 +18,7 @@ use api::v1::greptime_database_client::GreptimeDatabaseClient;
 use api::v1::greptime_request::Request;
 use api::v1::query_request::Query;
 use api::v1::{
-    AlterExpr, AuthHeader, CreateTableExpr, DdlRequest, GreptimeRequest, InsertRequests,
+    AlterTableExpr, AuthHeader, CreateTableExpr, DdlRequest, GreptimeRequest, InsertRequests,
     QueryRequest, RequestHeader,
 };
 use arrow_flight::Ticket;
@@ -211,9 +211,9 @@ impl Database {
         .await
     }
 
-    pub async fn alter(&self, expr: AlterExpr) -> Result<Output> {
+    pub async fn alter(&self, expr: AlterTableExpr) -> Result<Output> {
         self.do_get(Request::Ddl(DdlRequest {
-            expr: Some(DdlExpr::Alter(expr)),
+            expr: Some(DdlExpr::AlterTable(expr)),
         }))
         .await
     }

src/cmd/Cargo.toml

@@ -10,7 +10,7 @@ name = "greptime"
 path = "src/bin/greptime.rs"
 
 [features]
-default = ["python"]
+default = ["python", "servers/pprof", "servers/mem-prof"]
 tokio-console = ["common-telemetry/tokio-console"]
 python = ["frontend/python"]
 
@@ -25,6 +25,7 @@ cache.workspace = true
 catalog.workspace = true
 chrono.workspace = true
 clap.workspace = true
+cli.workspace = true
 client.workspace = true
 common-base.workspace = true
 common-catalog.workspace = true
@@ -33,6 +34,7 @@ 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
@@ -52,6 +54,7 @@ flow.workspace = true
 frontend = { workspace = true, default-features = false }
 futures.workspace = true
 human-panic = "2.0"
+humantime.workspace = true
 lazy_static.workspace = true
 meta-client.workspace = true
 meta-srv.workspace = true
@@ -78,7 +81,7 @@ table.workspace = true
 tokio.workspace = true
 toml.workspace = true
 tonic.workspace = true
-tracing-appender = "0.2"
+tracing-appender.workspace = true
 
 [target.'cfg(not(windows))'.dependencies]
 tikv-jemallocator = "0.6"

src/cmd/src/bin/greptime.rs

@@ -15,10 +15,11 @@
 #![doc = include_str!("../../../../README.md")]
 
 use clap::{Parser, Subcommand};
-use cmd::error::Result;
+use cmd::error::{InitTlsProviderSnafu, Result};
 use cmd::options::GlobalOptions;
 use cmd::{cli, datanode, flownode, frontend, metasrv, standalone, App};
 use common_version::version;
+use servers::install_ring_crypto_provider;
 
 #[derive(Parser)]
 #[command(name = "greptime", author, version, long_version = version(), about)]
@@ -94,6 +95,7 @@ async fn main() -> Result<()> {
 
 async fn main_body() -> Result<()> {
     setup_human_panic();
+    install_ring_crypto_provider().map_err(|msg| InitTlsProviderSnafu { msg }.build())?;
     start(Command::parse()).await
 }
 

src/cmd/src/cli.rs

@@ -12,39 +12,17 @@
 // See the License for the specific language governing permissions and
 // limitations under the License.
 
-mod bench;
-
-// Wait for https://github.com/GreptimeTeam/greptimedb/issues/2373
-#[allow(unused)]
-mod cmd;
-mod export;
-mod helper;
-
-// Wait for https://github.com/GreptimeTeam/greptimedb/issues/2373
-mod database;
-mod import;
-#[allow(unused)]
-mod repl;
-
-use async_trait::async_trait;
-use bench::BenchTableMetadataCommand;
 use clap::Parser;
+use cli::Tool;
 use common_telemetry::logging::{LoggingOptions, TracingOptions};
-pub use repl::Repl;
+use plugins::SubCommand;
+use snafu::ResultExt;
 use tracing_appender::non_blocking::WorkerGuard;
 
-use self::export::ExportCommand;
-use crate::cli::import::ImportCommand;
-use crate::error::Result;
 use crate::options::GlobalOptions;
-use crate::App;
-
+use crate::{error, App, Result};
 pub const APP_NAME: &str = "greptime-cli";
-
-#[async_trait]
-pub trait Tool: Send + Sync {
-    async fn do_work(&self) -> Result<()>;
-}
+use async_trait::async_trait;
 
 pub struct Instance {
     tool: Box<dyn Tool>,
@@ -54,12 +32,16 @@ pub struct Instance {
 }
 
 impl Instance {
-    fn new(tool: Box<dyn Tool>, guard: Vec<WorkerGuard>) -> Self {
+    pub fn new(tool: Box<dyn Tool>, guard: Vec<WorkerGuard>) -> Self {
         Self {
             tool,
             _guard: guard,
         }
     }
+
+    pub async fn start(&mut self) -> Result<()> {
+        self.tool.do_work().await.context(error::StartCliSnafu)
+    }
 }
 
 #[async_trait]
@@ -69,7 +51,8 @@ impl App for Instance {
     }
 
     async fn start(&mut self) -> Result<()> {
-        self.tool.do_work().await
+        self.start().await.unwrap();
+        Ok(())
     }
 
     fn wait_signal(&self) -> bool {
@@ -96,7 +79,12 @@ impl Command {
             None,
         );
 
-        self.cmd.build(guard).await
+        let tool = self.cmd.build().await.context(error::BuildCliSnafu)?;
+        let instance = Instance {
+            tool,
+            _guard: guard,
+        };
+        Ok(instance)
     }
 
     pub fn load_options(&self, global_options: &GlobalOptions) -> Result<LoggingOptions> {
@@ -112,38 +100,81 @@ impl Command {
     }
 }
 
-#[derive(Parser)]
-enum SubCommand {
-    // Attach(AttachCommand),
-    Bench(BenchTableMetadataCommand),
-    Export(ExportCommand),
-    Import(ImportCommand),
-}
+#[cfg(test)]
+mod tests {
+    use clap::Parser;
+    use client::{Client, Database};
+    use common_catalog::consts::{DEFAULT_CATALOG_NAME, DEFAULT_SCHEMA_NAME};
+    use common_telemetry::logging::LoggingOptions;
 
-impl SubCommand {
-    async fn build(&self, guard: Vec<WorkerGuard>) -> Result<Instance> {
-        match self {
-            // SubCommand::Attach(cmd) => cmd.build().await,
-            SubCommand::Bench(cmd) => cmd.build(guard).await,
-            SubCommand::Export(cmd) => cmd.build(guard).await,
-            SubCommand::Import(cmd) => cmd.build(guard).await,
-        }
-    }
-}
-
-#[derive(Debug, Parser)]
-pub(crate) struct AttachCommand {
-    #[clap(long)]
-    pub(crate) grpc_addr: String,
-    #[clap(long)]
-    pub(crate) meta_addr: Option<String>,
-    #[clap(long, action)]
-    pub(crate) disable_helper: bool,
-}
-
-impl AttachCommand {
-    #[allow(dead_code)]
-    async fn build(self) -> Result<Instance> {
-        unimplemented!("Wait for https://github.com/GreptimeTeam/greptimedb/issues/2373")
+    use crate::error::Result as CmdResult;
+    use crate::options::GlobalOptions;
+    use crate::{cli, standalone, App};
+
+    #[tokio::test(flavor = "multi_thread")]
+    async fn test_export_create_table_with_quoted_names() -> CmdResult<()> {
+        let output_dir = tempfile::tempdir().unwrap();
+
+        let standalone = standalone::Command::parse_from([
+            "standalone",
+            "start",
+            "--data-home",
+            &*output_dir.path().to_string_lossy(),
+        ]);
+
+        let standalone_opts = standalone.load_options(&GlobalOptions::default()).unwrap();
+        let mut instance = standalone.build(standalone_opts).await?;
+        instance.start().await?;
+
+        let client = Client::with_urls(["127.0.0.1:4001"]);
+        let database = Database::new(DEFAULT_CATALOG_NAME, DEFAULT_SCHEMA_NAME, client);
+        database
+            .sql(r#"CREATE DATABASE "cli.export.create_table";"#)
+            .await
+            .unwrap();
+        database
+            .sql(
+                r#"CREATE TABLE "cli.export.create_table"."a.b.c"(
+                        ts TIMESTAMP,
+                        TIME INDEX (ts)
+                    ) engine=mito;
+                "#,
+            )
+            .await
+            .unwrap();
+
+        let output_dir = tempfile::tempdir().unwrap();
+        let cli = cli::Command::parse_from([
+            "cli",
+            "export",
+            "--addr",
+            "127.0.0.1:4000",
+            "--output-dir",
+            &*output_dir.path().to_string_lossy(),
+            "--target",
+            "schema",
+        ]);
+        let mut cli_app = cli.build(LoggingOptions::default()).await?;
+        cli_app.start().await?;
+
+        instance.stop().await?;
+
+        let output_file = output_dir
+            .path()
+            .join("greptime")
+            .join("cli.export.create_table")
+            .join("create_tables.sql");
+        let res = std::fs::read_to_string(output_file).unwrap();
+        let expect = r#"CREATE TABLE IF NOT EXISTS "a.b.c" (
+  "ts" TIMESTAMP(3) NOT NULL,
+  TIME INDEX ("ts")
+)
+
+ENGINE=mito
+;
+"#;
+        assert_eq!(res.trim(), expect.trim());
+
+        Ok(())
     }
 }

src/cmd/src/datanode.rs

@@ -16,10 +16,12 @@ use std::sync::Arc;
 use std::time::Duration;
 
 use async_trait::async_trait;
+use cache::build_datanode_cache_registry;
 use catalog::kvbackend::MetaKvBackend;
 use clap::Parser;
 use common_base::Plugins;
 use common_config::Configurable;
+use common_meta::cache::LayeredCacheRegistryBuilder;
 use common_telemetry::logging::TracingOptions;
 use common_telemetry::{info, warn};
 use common_version::{short_version, version};
@@ -57,10 +59,6 @@ impl Instance {
         }
     }
 
-    pub fn datanode_mut(&mut self) -> &mut Datanode {
-        &mut self.datanode
-    }
-
     pub fn datanode(&self) -> &Datanode {
         &self.datanode
     }
@@ -272,9 +270,10 @@ impl StartCommand {
         info!("Datanode start command: {:#?}", self);
         info!("Datanode options: {:#?}", opts);
 
+        let plugin_opts = opts.plugins;
         let opts = opts.component;
         let mut plugins = Plugins::new();
-        plugins::setup_datanode_plugins(&mut plugins, &opts)
+        plugins::setup_datanode_plugins(&mut plugins, &plugin_opts, &opts)
             .await
             .context(StartDatanodeSnafu)?;
 
@@ -299,9 +298,17 @@ impl StartCommand {
             client: meta_client.clone(),
         });
 
+        // Builds cache registry for datanode.
+        let layered_cache_registry = Arc::new(
+            LayeredCacheRegistryBuilder::default()
+                .add_cache_registry(build_datanode_cache_registry(meta_backend.clone()))
+                .build(),
+        );
+
         let mut datanode = DatanodeBuilder::new(opts.clone(), plugins)
             .with_meta_client(meta_client)
             .with_kv_backend(meta_backend)
+            .with_cache_registry(layered_cache_registry)
             .build()
             .await
             .context(StartDatanodeSnafu)?;

src/cmd/src/error.rs

@@ -24,6 +24,12 @@ use snafu::{Location, 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)]
@@ -108,6 +114,20 @@ pub enum Error {
         source: frontend::error::Error,
     },
 
+    #[snafu(display("Failed to build cli"))]
+    BuildCli {
+        #[snafu(implicit)]
+        location: Location,
+        source: BoxedError,
+    },
+
+    #[snafu(display("Failed to start cli"))]
+    StartCli {
+        #[snafu(implicit)]
+        location: Location,
+        source: BoxedError,
+    },
+
     #[snafu(display("Failed to build meta server"))]
     BuildMetaServer {
         #[snafu(implicit)]
@@ -340,6 +360,8 @@ impl ErrorExt for Error {
             Error::ShutdownMetaServer { source, .. } => source.status_code(),
             Error::BuildMetaServer { source, .. } => source.status_code(),
             Error::UnsupportedSelectorType { source, .. } => source.status_code(),
+            Error::BuildCli { source, .. } => source.status_code(),
+            Error::StartCli { source, .. } => source.status_code(),
 
             Error::InitMetadata { source, .. } | Error::InitDdlManager { source, .. } => {
                 source.status_code()
@@ -369,9 +391,10 @@ impl ErrorExt for Error {
             }
             Error::SubstraitEncodeLogicalPlan { source, .. } => source.status_code(),
 
-            Error::SerdeJson { .. } | Error::FileIo { .. } | Error::SpawnThread { .. } => {
-                StatusCode::Unexpected
-            }
+            Error::SerdeJson { .. }
+            | Error::FileIo { .. }
+            | Error::SpawnThread { .. }
+            | Error::InitTlsProvider { .. } => StatusCode::Unexpected,
 
             Error::Other { source, .. } => source.status_code(),
 

src/cmd/src/flownode.rs

@@ -15,13 +15,15 @@
 use std::sync::Arc;
 
 use cache::{build_fundamental_cache_registry, with_default_composite_cache_registry};
-use catalog::kvbackend::{CachedMetaKvBackendBuilder, KvBackendCatalogManager, MetaKvBackend};
+use catalog::information_extension::DistributedInformationExtension;
+use catalog::kvbackend::{CachedKvBackendBuilder, KvBackendCatalogManager, MetaKvBackend};
 use clap::Parser;
 use client::client_manager::NodeClients;
 use common_base::Plugins;
 use common_config::Configurable;
 use common_grpc::channel_manager::ChannelConfig;
 use common_meta::cache::{CacheRegistryBuilder, LayeredCacheRegistryBuilder};
+use common_meta::heartbeat::handler::invalidate_table_cache::InvalidateCacheHandler;
 use common_meta::heartbeat::handler::parse_mailbox_message::ParseMailboxMessageHandler;
 use common_meta::heartbeat::handler::HandlerGroupExecutor;
 use common_meta::key::flow::FlowMetadataManager;
@@ -30,7 +32,6 @@ use common_telemetry::info;
 use common_telemetry::logging::TracingOptions;
 use common_version::{short_version, version};
 use flow::{FlownodeBuilder, FlownodeInstance, FrontendInvoker};
-use frontend::heartbeat::handler::invalidate_table_cache::InvalidateTableCacheHandler;
 use meta_client::{MetaClientOptions, MetaClientType};
 use servers::Mode;
 use snafu::{OptionExt, ResultExt};
@@ -62,10 +63,6 @@ impl Instance {
         }
     }
 
-    pub fn flownode_mut(&mut self) -> &mut FlownodeInstance {
-        &mut self.flownode
-    }
-
     pub fn flownode(&self) -> &FlownodeInstance {
         &self.flownode
     }
@@ -246,11 +243,12 @@ impl StartCommand {
         let cache_tti = meta_config.metadata_cache_tti;
 
         // TODO(discord9): add helper function to ease the creation of cache registry&such
-        let cached_meta_backend = CachedMetaKvBackendBuilder::new(meta_client.clone())
-            .cache_max_capacity(cache_max_capacity)
-            .cache_ttl(cache_ttl)
-            .cache_tti(cache_tti)
-            .build();
+        let cached_meta_backend =
+            CachedKvBackendBuilder::new(Arc::new(MetaKvBackend::new(meta_client.clone())))
+                .cache_max_capacity(cache_max_capacity)
+                .cache_ttl(cache_ttl)
+                .cache_tti(cache_tti)
+                .build();
         let cached_meta_backend = Arc::new(cached_meta_backend);
 
         // Builds cache registry
@@ -269,11 +267,13 @@ impl StartCommand {
             .build(),
         );
 
+        let information_extension =
+            Arc::new(DistributedInformationExtension::new(meta_client.clone()));
         let catalog_manager = KvBackendCatalogManager::new(
-            opts.mode,
-            Some(meta_client.clone()),
+            information_extension,
             cached_meta_backend.clone(),
             layered_cache_registry.clone(),
+            None,
         );
 
         let table_metadata_manager =
@@ -285,9 +285,7 @@ impl StartCommand {
 
         let executor = HandlerGroupExecutor::new(vec![
             Arc::new(ParseMailboxMessageHandler),
-            Arc::new(InvalidateTableCacheHandler::new(
-                layered_cache_registry.clone(),
-            )),
+            Arc::new(InvalidateCacheHandler::new(layered_cache_registry.clone())),
         ]);
 
         let heartbeat_task = flow::heartbeat::HeartbeatTask::new(