chore: upgrade 0.4.2 (#2644)

* fix: prune predicate show use real schema to create physical exprs

* refactor: remove redundant results

* fix: unit tests

* test: add more sqlness cases

* test: add more sqlness cases

* fix: sqlness orderby

* chore: update log

* fix: cache physical expr in memtable iter

---------

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

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

@@ -1,93 +0,0 @@
-name: Build and push dev-builder image
-description: Build and push dev-builder image to DockerHub and ACR
-inputs:
-  dockerhub-image-registry:
-    description: The dockerhub image registry to store the images
-    required: false
-    default: docker.io
-  dockerhub-image-registry-username:
-    description: The dockerhub username to login to the image registry
-    required: true
-  dockerhub-image-registry-token:
-    description: The dockerhub token to login to the image registry
-    required: true
-  dockerhub-image-namespace:
-    description: The dockerhub namespace of the image registry to store the images
-    required: false
-    default: greptime
-  acr-image-registry:
-    description: The ACR image registry to store the images
-    required: true
-  acr-image-registry-username:
-    description: The ACR username to login to the image registry
-    required: true
-  acr-image-registry-password:
-    description: The ACR password to login to the image registry
-    required: true
-  acr-image-namespace:
-    description: The ACR namespace of the image registry to store the images
-    required: false
-    default: greptime
-  version:
-    description: Version of the dev-builder
-    required: false
-    default: latest
-runs:
-  using: composite
-  steps:
-    - name: Login to Dockerhub
-      uses: docker/login-action@v2
-      with:
-        registry: ${{ inputs.dockerhub-image-registry }}
-        username: ${{ inputs.dockerhub-image-registry-username }}
-        password: ${{ inputs.dockerhub-image-registry-token }}
-
-    - name: Build and push ubuntu dev builder image to dockerhub
-      shell: bash
-      run:
-        make dev-builder \
-        BASE_IMAGE=ubuntu \
-        BUILDX_MULTI_PLATFORM_BUILD=true \
-        IMAGE_REGISTRY=${{ inputs.dockerhub-image-registry }} \
-        IMAGE_NAMESPACE=${{ inputs.dockerhub-image-namespace }} \
-        IMAGE_TAG=${{ inputs.version }}
-
-    - name: Build and push centos dev builder image to dockerhub
-      shell: bash
-      run:
-        make dev-builder \
-        BASE_IMAGE=centos \
-        BUILDX_MULTI_PLATFORM_BUILD=true \
-        IMAGE_REGISTRY=${{ inputs.dockerhub-image-registry }} \
-        IMAGE_NAMESPACE=${{ inputs.dockerhub-image-namespace }} \
-        IMAGE_TAG=${{ inputs.version }}
-
-    - name: Login to ACR
-      uses: docker/login-action@v2
-      continue-on-error: true
-      with:
-        registry: ${{ inputs.acr-image-registry }}
-        username: ${{ inputs.acr-image-registry-username }}
-        password: ${{ inputs.acr-image-registry-password }}
-
-    - name: Build and push ubuntu dev builder image to ACR
-      shell: bash
-      continue-on-error: true
-      run: # buildx will cache the images that already built, so it will not take long time to build the images again.
-        make dev-builder \
-        BASE_IMAGE=ubuntu \
-        BUILDX_MULTI_PLATFORM_BUILD=true \
-        IMAGE_REGISTRY=${{ inputs.acr-image-registry }} \
-        IMAGE_NAMESPACE=${{ inputs.acr-image-namespace }} \
-        IMAGE_TAG=${{ inputs.version }}
-
-    - name: Build and push centos dev builder image to ACR
-      shell: bash
-      continue-on-error: true
-      run: # buildx will cache the images that already built, so it will not take long time to build the images again.
-        make dev-builder \
-        BASE_IMAGE=centos \
-        BUILDX_MULTI_PLATFORM_BUILD=true \
-        IMAGE_REGISTRY=${{ inputs.acr-image-registry }} \
-        IMAGE_NAMESPACE=${{ inputs.acr-image-namespace }} \
-        IMAGE_TAG=${{ inputs.version }}

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

@@ -0,0 +1,76 @@
+name: Build and push dev-builder images
+description: Build and push dev-builder images to DockerHub and ACR
+inputs:
+  dockerhub-image-registry:
+    description: The dockerhub image registry to store the images
+    required: false
+    default: docker.io
+  dockerhub-image-registry-username:
+    description: The dockerhub username to login to the image registry
+    required: true
+  dockerhub-image-registry-token:
+    description: The dockerhub token to login to the image registry
+    required: true
+  dockerhub-image-namespace:
+    description: The dockerhub namespace of the image registry to store the images
+    required: false
+    default: greptime
+  version:
+    description: Version of the dev-builder
+    required: false
+    default: latest
+  build-dev-builder-ubuntu:
+    description: Build dev-builder-ubuntu image
+    required: false
+    default: 'true'
+  build-dev-builder-centos:
+    description: Build dev-builder-centos image
+    required: false
+    default: 'true'
+  build-dev-builder-android:
+    description: Build dev-builder-android image
+    required: false
+    default: 'true'
+runs:
+  using: composite
+  steps:
+    - name: Login to Dockerhub
+      uses: docker/login-action@v2
+      with:
+        registry: ${{ inputs.dockerhub-image-registry }}
+        username: ${{ inputs.dockerhub-image-registry-username }}
+        password: ${{ inputs.dockerhub-image-registry-token }}
+
+    - name: Build and push dev-builder-ubuntu image
+      shell: bash
+      if: ${{ inputs.build-dev-builder-ubuntu == 'true' }}
+      run: |
+        make dev-builder \
+          BASE_IMAGE=ubuntu \
+          BUILDX_MULTI_PLATFORM_BUILD=true \
+          IMAGE_REGISTRY=${{ inputs.dockerhub-image-registry }} \
+          IMAGE_NAMESPACE=${{ inputs.dockerhub-image-namespace }} \
+          IMAGE_TAG=${{ inputs.version }}
+
+    - name: Build and push dev-builder-centos image
+      shell: bash
+      if: ${{ inputs.build-dev-builder-centos == 'true' }}
+      run: |
+        make dev-builder \
+          BASE_IMAGE=centos \
+          BUILDX_MULTI_PLATFORM_BUILD=true \
+          IMAGE_REGISTRY=${{ inputs.dockerhub-image-registry }} \
+          IMAGE_NAMESPACE=${{ inputs.dockerhub-image-namespace }} \
+          IMAGE_TAG=${{ inputs.version }}
+
+    - name: Build and push dev-builder-android image # Only build image for amd64 platform.
+      shell: bash
+      if: ${{ inputs.build-dev-builder-android == 'true' }}
+      run: |
+        make dev-builder \
+          BASE_IMAGE=android \
+          IMAGE_REGISTRY=${{ inputs.dockerhub-image-registry }} \
+          IMAGE_NAMESPACE=${{ inputs.dockerhub-image-namespace }} \
+          IMAGE_TAG=${{ inputs.version }} && \
+        
+        docker push ${{ inputs.dockerhub-image-registry }}/${{ inputs.dockerhub-image-namespace }}/dev-builder-android:${{ inputs.version }}

.github/actions/build-greptime-binary/action.yml

@@ -16,35 +16,20 @@ inputs:
   version:
     description: Version of the artifact
     required: true
-  release-to-s3-bucket:
-    description: S3 bucket to store released artifacts
-    required: true
-  aws-access-key-id:
-    description: AWS access key id
-    required: true
-  aws-secret-access-key:
-    description: AWS secret access key
-    required: true
-  aws-region:
-    description: AWS region
-    required: true
-  upload-to-s3:
-    description: Upload to S3
-    required: false
-    default: 'true'
-  upload-latest-artifacts:
-    description: Upload the latest artifacts to S3
-    required: false
-    default: 'true'
   working-dir:
     description: Working directory to build the artifacts
     required: false
     default: .
+  build-android-artifacts:
+    description: Build android artifacts
+    required: false
+    default: 'false'
 runs:
   using: composite
   steps:
     - name: Build greptime binary
       shell: bash
+      if: ${{ inputs.build-android-artifacts == 'false' }}
       run: |
         cd ${{ inputs.working-dir }} && \
         make build-by-dev-builder \
@@ -54,14 +39,25 @@ runs:
 
     - name: Upload artifacts
       uses: ./.github/actions/upload-artifacts
+      if: ${{ inputs.build-android-artifacts == 'false' }}
       with:
         artifacts-dir: ${{ inputs.artifacts-dir }}
         target-file: ./target/${{ inputs.cargo-profile }}/greptime
         version: ${{ inputs.version }}
-        release-to-s3-bucket: ${{ inputs.release-to-s3-bucket }}
-        aws-access-key-id: ${{ inputs.aws-access-key-id }}
-        aws-secret-access-key: ${{ inputs.aws-secret-access-key }}
-        aws-region: ${{ inputs.aws-region }}
-        upload-to-s3: ${{ inputs.upload-to-s3 }}
-        upload-latest-artifacts: ${{ inputs.upload-latest-artifacts }}
+        working-dir: ${{ inputs.working-dir }}
+
+    # TODO(zyy17): We can remove build-android-artifacts flag in the future.
+    - name: Build greptime binary
+      shell: bash
+      if: ${{ inputs.build-android-artifacts == 'true' }}
+      run: |
+        cd ${{ inputs.working-dir }} && make strip-android-bin
+
+    - name: Upload android artifacts
+      uses: ./.github/actions/upload-artifacts
+      if: ${{ inputs.build-android-artifacts == 'true' }}
+      with:
+        artifacts-dir: ${{ inputs.artifacts-dir }}
+        target-file: ./target/aarch64-linux-android/release/greptime
+        version: ${{ inputs.version }}
         working-dir: ${{ inputs.working-dir }}

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

@@ -13,30 +13,10 @@ inputs:
   disable-run-tests:
     description: Disable running integration tests
     required: true
-  release-to-s3-bucket:
-    description: S3 bucket to store released artifacts
-    required: true
-  aws-access-key-id:
-    description: AWS access key id
-    required: true
-  aws-secret-access-key:
-    description: AWS secret access key
-    required: true
-  aws-region:
-    description: AWS region
-    required: true
   dev-mode:
     description: Enable dev mode, only build standard greptime
     required: false
     default: 'false'
-  upload-to-s3:
-    description: Upload to S3
-    required: false
-    default: 'true'
-  upload-latest-artifacts:
-    description: Upload the latest artifacts to S3
-    required: false
-    default: 'true'
   working-dir:
     description: Working directory to build the artifacts
     required: false
@@ -68,12 +48,6 @@ runs:
         cargo-profile: ${{ inputs.cargo-profile }}
         artifacts-dir: greptime-linux-${{ inputs.arch }}-pyo3-${{ inputs.version }}
         version: ${{ inputs.version }}
-        release-to-s3-bucket: ${{ inputs.release-to-s3-bucket }}
-        aws-access-key-id: ${{ inputs.aws-access-key-id }}
-        aws-secret-access-key: ${{ inputs.aws-secret-access-key }}
-        aws-region: ${{ inputs.aws-region }}
-        upload-to-s3: ${{ inputs.upload-to-s3 }}
-        upload-latest-artifacts: ${{ inputs.upload-latest-artifacts }}
         working-dir: ${{ inputs.working-dir }}
 
     - name: Build greptime without pyo3
@@ -85,12 +59,6 @@ runs:
         cargo-profile: ${{ inputs.cargo-profile }}
         artifacts-dir: greptime-linux-${{ inputs.arch }}-${{ inputs.version }}
         version: ${{ inputs.version }}
-        release-to-s3-bucket: ${{ inputs.release-to-s3-bucket }}
-        aws-access-key-id: ${{ inputs.aws-access-key-id }}
-        aws-secret-access-key: ${{ inputs.aws-secret-access-key }}
-        aws-region: ${{ inputs.aws-region }}
-        upload-to-s3: ${{ inputs.upload-to-s3 }}
-        upload-latest-artifacts: ${{ inputs.upload-latest-artifacts }}
         working-dir: ${{ inputs.working-dir }}
 
     - name: Clean up the target directory # Clean up the target directory for the centos7 base image, or it will still use the objects of last build.
@@ -107,10 +75,14 @@ runs:
         cargo-profile: ${{ inputs.cargo-profile }}
         artifacts-dir: greptime-linux-${{ inputs.arch }}-centos-${{ inputs.version }}
         version: ${{ inputs.version }}
-        release-to-s3-bucket: ${{ inputs.release-to-s3-bucket }}
-        aws-access-key-id: ${{ inputs.aws-access-key-id }}
-        aws-secret-access-key: ${{ inputs.aws-secret-access-key }}
-        aws-region: ${{ inputs.aws-region }}
-        upload-to-s3: ${{ inputs.upload-to-s3 }}
-        upload-latest-artifacts: ${{ inputs.upload-latest-artifacts }}
         working-dir: ${{ inputs.working-dir }}
+
+    - name: Build greptime on android base image
+      uses: ./.github/actions/build-greptime-binary
+      if: ${{ inputs.arch == 'amd64' && inputs.dev-mode == 'false' }} # Only build android base image on amd64.
+      with:
+        base-image: android
+        artifacts-dir: greptime-android-arm64-${{ inputs.version }}
+        version: ${{ inputs.version }}
+        working-dir: ${{ inputs.working-dir }}
+        build-android-artifacts: true

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

@@ -19,25 +19,9 @@ inputs:
   disable-run-tests:
     description: Disable running integration tests
     required: true
-  release-to-s3-bucket:
-    description: S3 bucket to store released artifacts
-    required: true
   artifacts-dir:
     description: Directory to store artifacts
     required: true
-  aws-access-key-id:
-    description: AWS access key id
-    required: true
-  aws-secret-access-key:
-    description: AWS secret access key
-    required: true
-  aws-region:
-    description: AWS region
-    required: true
-  upload-to-s3:
-    description: Upload to S3
-    required: false
-    default: 'true'
 runs:
   using: composite
   steps:
@@ -103,8 +87,3 @@ runs:
         artifacts-dir: ${{ inputs.artifacts-dir }}
         target-file: target/${{ inputs.arch }}/${{ inputs.cargo-profile }}/greptime
         version: ${{ inputs.version }}
-        release-to-s3-bucket: ${{ inputs.release-to-s3-bucket }}
-        aws-access-key-id: ${{ inputs.aws-access-key-id }}
-        aws-secret-access-key: ${{ inputs.aws-secret-access-key }}
-        aws-region: ${{ inputs.aws-region }}
-        upload-to-s3: ${{ inputs.upload-to-s3 }}

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

@@ -0,0 +1,80 @@
+name: Build Windows artifacts
+description: Build Windows artifacts
+inputs:
+  arch:
+    description: Architecture to build
+    required: true
+  rust-toolchain:
+    description: Rust toolchain to use
+    required: true
+  cargo-profile:
+    description: Cargo profile to build
+    required: true
+  features:
+    description: Cargo features to build
+    required: true
+  version:
+    description: Version of the artifact
+    required: true
+  disable-run-tests:
+    description: Disable running integration tests
+    required: true
+  artifacts-dir:
+    description: Directory to store artifacts
+    required: true
+runs:
+  using: composite
+  steps:
+    - uses: arduino/setup-protoc@v1
+
+    - name: Install rust toolchain
+      uses: dtolnay/rust-toolchain@master
+      with:
+        toolchain: ${{ inputs.rust-toolchain }}
+        targets: ${{ inputs.arch }}
+        components: llvm-tools-preview
+
+    - name: Rust Cache
+      uses: Swatinem/rust-cache@v2
+
+    - name: Install Python
+      uses: actions/setup-python@v4
+      with:
+        python-version: '3.10'
+
+    - name: Install PyArrow Package
+      shell: pwsh
+      run: pip install pyarrow
+
+    - name: Install WSL distribution
+      uses: Vampire/setup-wsl@v2
+      with:
+        distribution: Ubuntu-22.04
+
+    - name: Install latest nextest release # For integration tests.
+      if: ${{ inputs.disable-run-tests == 'false' }}
+      uses: taiki-e/install-action@nextest
+
+    - name: Run integration tests
+      if: ${{ inputs.disable-run-tests == 'false' }}
+      shell: pwsh
+      run: make test sqlness-test
+
+    - name: Upload sqlness logs
+      if: ${{ failure() }} # Only upload logs when the integration tests failed.
+      uses: actions/upload-artifact@v3
+      with:
+        name: sqlness-logs
+        path: ${{ runner.temp }}/greptime-*.log
+        retention-days: 3
+
+    - name: Build greptime binary
+      shell: pwsh
+      run: cargo build --profile ${{ inputs.cargo-profile }} --features ${{ inputs.features }} --target ${{ inputs.arch }}
+
+    - name: Upload artifacts
+      uses: ./.github/actions/upload-artifacts
+      with:
+        artifacts-dir: ${{ inputs.artifacts-dir }}
+        target-file: target/${{ inputs.arch }}/${{ inputs.cargo-profile }}/greptime
+        version: ${{ inputs.version }}

.github/actions/publish-github-release/action.yml

@@ -1,5 +1,5 @@
-name: Release artifacts
-description: Release artifacts
+name: Publish GitHub release
+description: Publish GitHub release
 inputs:
   version:
     description: Version to release

.github/actions/release-cn-artifacts/action.yaml

@@ -0,0 +1,138 @@
+name: Release CN artifacts
+description: Release artifacts to CN region
+inputs:
+  src-image-registry:
+    description: The source image registry to store the images
+    required: true
+    default: docker.io
+  src-image-namespace:
+    description: The namespace of the source image registry to store the images
+    required: true
+    default: greptime
+  src-image-name:
+    description: The name of the source image
+    required: false
+    default: greptimedb
+  dst-image-registry:
+    description: The destination image registry to store the images
+    required: true
+  dst-image-namespace:
+    description: The namespace of the destination image registry to store the images
+    required: true
+    default: greptime
+  dst-image-registry-username:
+    description: The username to login to the image registry
+    required: true
+  dst-image-registry-password:
+    description: The password to login to the image registry
+    required: true
+  version:
+    description: Version of the artifact
+    required: true
+  dev-mode:
+    description: Enable dev mode, only push standard greptime
+    required: false
+    default: 'false'
+  push-latest-tag:
+    description: Whether to push the latest tag of the image
+    required: false
+    default: 'true'
+  aws-cn-s3-bucket:
+    description: S3 bucket to store released artifacts in CN region
+    required: true
+  aws-cn-access-key-id:
+    description: AWS access key id in CN region
+    required: true
+  aws-cn-secret-access-key:
+    description: AWS secret access key in CN region
+    required: true
+  aws-cn-region:
+    description: AWS region in CN
+    required: true
+  upload-to-s3:
+    description: Upload to S3
+    required: false
+    default: 'true'
+  artifacts-dir:
+    description: Directory to store artifacts
+    required: false
+    default: 'artifacts'
+  update-version-info:
+    description: Update the version info in S3
+    required: false
+    default: 'true'
+  upload-max-retry-times:
+    description: Max retry times for uploading artifacts to S3
+    required: false
+    default: "20"
+  upload-retry-timeout:
+    description: Timeout for uploading artifacts to S3
+    required: false
+    default: "30" # minutes
+runs:
+  using: composite
+  steps:
+    - name: Download artifacts
+      uses: actions/download-artifact@v3
+      with:
+        path: ${{ inputs.artifacts-dir }}
+
+    - name: Release artifacts to cn region
+      uses: nick-invision/retry@v2
+      if: ${{ inputs.upload-to-s3 == 'true' }}
+      env:
+        AWS_ACCESS_KEY_ID: ${{ inputs.aws-cn-access-key-id }}
+        AWS_SECRET_ACCESS_KEY: ${{ inputs.aws-cn-secret-access-key }}
+        AWS_DEFAULT_REGION: ${{ inputs.aws-cn-region }}
+        UPDATE_VERSION_INFO: ${{ inputs.update-version-info }}
+      with:
+        max_attempts: ${{ inputs.upload-max-retry-times }}
+        timeout_minutes: ${{ inputs.upload-retry-timeout }}
+        command: |
+          ./.github/scripts/upload-artifacts-to-s3.sh \
+            ${{ inputs.artifacts-dir }} \
+            ${{ inputs.version }} \
+            ${{ inputs.aws-cn-s3-bucket }}
+
+    - name: Push greptimedb image from Dockerhub to ACR
+      shell: bash
+      env:
+        DST_REGISTRY_USERNAME: ${{ inputs.dst-image-registry-username }}
+        DST_REGISTRY_PASSWORD: ${{ inputs.dst-image-registry-password }}
+      run: |
+        ./.github/scripts/copy-image.sh \
+         ${{ inputs.src-image-registry }}/${{ inputs.src-image-namespace }}/${{ inputs.src-image-name }}:${{ inputs.version }} \
+         ${{ inputs.dst-image-registry }}/${{ inputs.dst-image-namespace }}
+
+    - name: Push latest greptimedb image from Dockerhub to ACR
+      shell: bash
+      if: ${{ inputs.push-latest-tag == 'true' }}
+      env:
+        DST_REGISTRY_USERNAME: ${{ inputs.dst-image-registry-username }}
+        DST_REGISTRY_PASSWORD: ${{ inputs.dst-image-registry-password }}
+      run: |
+        ./.github/scripts/copy-image.sh \
+         ${{ inputs.src-image-registry }}/${{ inputs.src-image-namespace }}/${{ inputs.src-image-name }}:latest \
+         ${{ inputs.dst-image-registry }}/${{ inputs.dst-image-namespace }}
+
+    - name: Push greptimedb-centos image from DockerHub to ACR
+      shell: bash
+      if: ${{ inputs.dev-mode == 'false' }}
+      env:
+        DST_REGISTRY_USERNAME: ${{ inputs.dst-image-registry-username }}
+        DST_REGISTRY_PASSWORD: ${{ inputs.dst-image-registry-password }}
+      run: |
+        ./.github/scripts/copy-image.sh \
+         ${{ inputs.src-image-registry }}/${{ inputs.src-image-namespace }}/${{ inputs.src-image-name }}-centos:latest \
+         ${{ inputs.dst-image-registry }}/${{ inputs.dst-image-namespace }}
+
+    - name: Push greptimedb-centos image from DockerHub to ACR
+      shell: bash
+      if: ${{ inputs.dev-mode == 'false' && inputs.push-latest-tag == 'true' }}
+      env:
+        DST_REGISTRY_USERNAME: ${{ inputs.dst-image-registry-username }}
+        DST_REGISTRY_PASSWORD: ${{ inputs.dst-image-registry-password }}
+      run: |
+        ./.github/scripts/copy-image.sh \
+         ${{ inputs.src-image-registry }}/${{ inputs.src-image-namespace }}/${{ inputs.src-image-name }}-centos:latest \
+         ${{ inputs.dst-image-registry }}/${{ inputs.dst-image-namespace }}

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

@@ -10,34 +10,6 @@ inputs:
   version:
     description: Version of the artifact
     required: true
-  release-to-s3-bucket:
-    description: S3 bucket to store released artifacts
-    required: true
-  aws-access-key-id:
-    description: AWS access key id
-    required: true
-  aws-secret-access-key:
-    description: AWS secret access key
-    required: true
-  aws-region:
-    description: AWS region
-    required: true
-  upload-to-s3:
-    description: Upload to S3
-    required: false
-    default: 'true'
-  upload-latest-artifacts:
-    description: Upload the latest artifacts to S3
-    required: false
-    default: 'true'
-  upload-max-retry-times:
-    description: Max retry times for uploading artifacts to S3
-    required: false
-    default: "20"
-  upload-retry-timeout:
-    description: Timeout for uploading artifacts to S3
-    required: false
-    default: "30" # minutes
   working-dir:
     description: Working directory to upload the artifacts
     required: false
@@ -61,9 +33,21 @@ runs:
       working-directory: ${{ inputs.working-dir }}
       shell: bash
       run: |
-        tar -zcvf ${{ inputs.artifacts-dir }}.tar.gz ${{ inputs.artifacts-dir }} && \
+        tar -zcvf ${{ inputs.artifacts-dir }}.tar.gz ${{ inputs.artifacts-dir }}
+
+    - name: Calculate checksum
+      if: runner.os != 'Windows'
+      working-directory: ${{ inputs.working-dir }}
+      shell: bash
+      run: |
         echo $(shasum -a 256 ${{ inputs.artifacts-dir }}.tar.gz | cut -f1 -d' ') > ${{ inputs.artifacts-dir }}.sha256sum
 
+    - name: Calculate checksum on Windows
+      if: runner.os == 'Windows'
+      working-directory: ${{ inputs.working-dir }}
+      shell: pwsh
+      run: Get-FileHash ${{ inputs.artifacts-dir }}.tar.gz -Algorithm SHA256 | select -ExpandProperty Hash > ${{ inputs.artifacts-dir }}.sha256sum
+
     # Note: The artifacts will be double zip compressed(related issue: https://github.com/actions/upload-artifact/issues/39).
     # However, when we use 'actions/download-artifact@v3' to download the artifacts, it will be automatically unzipped.
     - name: Upload artifacts
@@ -77,49 +61,3 @@ runs:
       with:
         name: ${{ inputs.artifacts-dir }}.sha256sum
         path: ${{ inputs.working-dir }}/${{ inputs.artifacts-dir }}.sha256sum
-
-    - name: Upload artifacts to S3
-      if: ${{ inputs.upload-to-s3 == 'true' }}
-      uses: nick-invision/retry@v2
-      env:
-        AWS_ACCESS_KEY_ID: ${{ inputs.aws-access-key-id }}
-        AWS_SECRET_ACCESS_KEY: ${{ inputs.aws-secret-access-key }}
-        AWS_DEFAULT_REGION: ${{ inputs.aws-region }}
-      with:
-        max_attempts: ${{ inputs.upload-max-retry-times }}
-        timeout_minutes: ${{ inputs.upload-retry-timeout }}
-        # The bucket layout will be:
-        # releases/greptimedb
-        # ├── v0.1.0
-        # │   ├── greptime-darwin-amd64-pyo3-v0.1.0.sha256sum
-        # │   └── greptime-darwin-amd64-pyo3-v0.1.0.tar.gz
-        # └── v0.2.0
-        #    ├── greptime-darwin-amd64-pyo3-v0.2.0.sha256sum
-        #    └── greptime-darwin-amd64-pyo3-v0.2.0.tar.gz
-        command: |
-          cd ${{ inputs.working-dir }} && \
-          aws s3 cp \
-            ${{ inputs.artifacts-dir }}.tar.gz \
-            s3://${{ inputs.release-to-s3-bucket }}/releases/greptimedb/${{ inputs.version }}/${{ inputs.artifacts-dir }}.tar.gz && \
-          aws s3 cp \
-            ${{ inputs.artifacts-dir }}.sha256sum \
-            s3://${{ inputs.release-to-s3-bucket }}/releases/greptimedb/${{ inputs.version }}/${{ inputs.artifacts-dir }}.sha256sum
-
-    - name: Upload latest artifacts to S3
-      if: ${{ inputs.upload-to-s3 == 'true' && inputs.upload-latest-artifacts == 'true' }} # We'll also upload the latest artifacts to S3 in the scheduled and formal release.
-      uses: nick-invision/retry@v2
-      env:
-        AWS_ACCESS_KEY_ID: ${{ inputs.aws-access-key-id }}
-        AWS_SECRET_ACCESS_KEY: ${{ inputs.aws-secret-access-key }}
-        AWS_DEFAULT_REGION: ${{ inputs.aws-region }}
-      with:
-        max_attempts: ${{ inputs.upload-max-retry-times }}
-        timeout_minutes: ${{ inputs.upload-retry-timeout }}
-        command: |
-          cd ${{ inputs.working-dir }} && \
-          aws s3 cp \
-            ${{ inputs.artifacts-dir }}.tar.gz \
-            s3://${{ inputs.release-to-s3-bucket }}/releases/greptimedb/latest/${{ inputs.artifacts-dir }}.tar.gz && \
-          aws s3 cp \
-            ${{ inputs.artifacts-dir }}.sha256sum \
-            s3://${{ inputs.release-to-s3-bucket }}/releases/greptimedb/latest/${{ inputs.artifacts-dir }}.sha256sum

.github/scripts/copy-image.sh

@@ -0,0 +1,47 @@
+#!/usr/bin/env bash
+
+set -e
+set -o pipefail
+
+SRC_IMAGE=$1
+DST_REGISTRY=$2
+SKOPEO_STABLE_IMAGE="quay.io/skopeo/stable:latest"
+
+# Check if necessary variables are set.
+function check_vars() {
+  for var in DST_REGISTRY_USERNAME DST_REGISTRY_PASSWORD DST_REGISTRY SRC_IMAGE; do
+    if [ -z "${!var}" ]; then
+      echo "$var is not set or empty."
+      echo "Usage: DST_REGISTRY_USERNAME=<your-dst-registry-username> DST_REGISTRY_PASSWORD=<your-dst-registry-password> $0 <dst-registry> <src-image>"
+      exit 1
+    fi
+  done
+}
+
+# Copies images from DockerHub to the destination registry.
+function copy_images_from_dockerhub() {
+  # Check if docker is installed.
+  if ! command -v docker &> /dev/null; then
+    echo "docker is not installed. Please install docker to continue."
+    exit 1
+  fi
+
+  # Extract the name and tag of the source image.
+  IMAGE_NAME=$(echo "$SRC_IMAGE" | sed "s/.*\///")
+
+  echo "Copying $SRC_IMAGE to $DST_REGISTRY/$IMAGE_NAME"
+
+  docker run "$SKOPEO_STABLE_IMAGE" copy -a docker://"$SRC_IMAGE" \
+    --dest-creds "$DST_REGISTRY_USERNAME":"$DST_REGISTRY_PASSWORD" \
+    docker://"$DST_REGISTRY/$IMAGE_NAME"
+}
+
+function main() {
+  check_vars
+  copy_images_from_dockerhub
+}
+
+# Usage example:
+# DST_REGISTRY_USERNAME=123 DST_REGISTRY_PASSWORD=456 \
+#   ./copy-image.sh greptime/greptimedb:v0.4.0 greptime-registry.cn-hangzhou.cr.aliyuncs.com
+main

.github/scripts/upload-artifacts-to-s3.sh

@@ -0,0 +1,102 @@
+#!/usr/bin/env bash
+
+set -e
+set -o pipefail
+
+ARTIFACTS_DIR=$1
+VERSION=$2
+AWS_S3_BUCKET=$3
+RELEASE_DIRS="releases/greptimedb"
+GREPTIMEDB_REPO="GreptimeTeam/greptimedb"
+
+# Check if necessary variables are set.
+function check_vars() {
+  for var in AWS_S3_BUCKET VERSION ARTIFACTS_DIR; do
+    if [ -z "${!var}" ]; then
+      echo "$var is not set or empty."
+      echo "Usage: $0 <artifacts-dir> <version> <aws-s3-bucket>"
+      exit 1
+    fi
+  done
+}
+
+# Uploads artifacts to AWS S3 bucket.
+function upload_artifacts() {
+  # The bucket layout will be:
+  # releases/greptimedb
+  # ├── latest-version.txt
+  # ├── latest-nightly-version.txt
+  # ├── v0.1.0
+  # │   ├── greptime-darwin-amd64-pyo3-v0.1.0.sha256sum
+  # │   └── greptime-darwin-amd64-pyo3-v0.1.0.tar.gz
+  # └── v0.2.0
+  #    ├── greptime-darwin-amd64-pyo3-v0.2.0.sha256sum
+  #    └── greptime-darwin-amd64-pyo3-v0.2.0.tar.gz
+  find "$ARTIFACTS_DIR" -type f \( -name "*.tar.gz" -o -name "*.sha256sum" \) | while IFS= read -r file; do
+    aws s3 cp \
+      "$file" "s3://$AWS_S3_BUCKET/$RELEASE_DIRS/$VERSION/$(basename "$file")"
+  done
+}
+
+# Updates the latest version information in AWS S3 if UPDATE_VERSION_INFO is true.
+function update_version_info() {
+  if [ "$UPDATE_VERSION_INFO" == "true" ]; then
+    # If it's the officail release(like v1.0.0, v1.0.1, v1.0.2, etc.), update latest-version.txt.
+    if [[ "$VERSION" =~ ^v[0-9]+\.[0-9]+\.[0-9]+$ ]]; then
+      echo "Updating latest-version.txt"
+      echo "$VERSION" > latest-version.txt
+      aws s3 cp \
+        latest-version.txt "s3://$AWS_S3_BUCKET/$RELEASE_DIRS/latest-version.txt"
+    fi
+
+    # If it's the nightly release, update latest-nightly-version.txt.
+    if [[ "$VERSION" == *"nightly"* ]]; then
+      echo "Updating latest-nightly-version.txt"
+      echo "$VERSION" > latest-nightly-version.txt
+      aws s3 cp \
+        latest-nightly-version.txt "s3://$AWS_S3_BUCKET/$RELEASE_DIRS/latest-nightly-version.txt"
+    fi
+  fi
+}
+
+# Downloads artifacts from Github if DOWNLOAD_ARTIFACTS_FROM_GITHUB is true.
+function download_artifacts_from_github() {
+  if [ "$DOWNLOAD_ARTIFACTS_FROM_GITHUB" == "true" ]; then
+    # Check if jq is installed.
+    if ! command -v jq &> /dev/null; then
+      echo "jq is not installed. Please install jq to continue."
+      exit 1
+    fi
+
+    # Get the latest release API response.
+    RELEASES_API_RESPONSE=$(curl -s -H "Accept: application/vnd.github.v3+json" "https://api.github.com/repos/$GREPTIMEDB_REPO/releases/latest")
+
+    # Extract download URLs for the artifacts.
+    # Exclude source code archives which are typically named as 'greptimedb-<version>.zip' or 'greptimedb-<version>.tar.gz'.
+    ASSET_URLS=$(echo "$RELEASES_API_RESPONSE" | jq -r '.assets[] | select(.name | test("greptimedb-.*\\.(zip|tar\\.gz)$") | not) | .browser_download_url')
+
+    # Download each asset.
+    while IFS= read -r url; do
+      if [ -n "$url" ]; then
+        curl -LJO "$url"
+        echo "Downloaded: $url"
+      fi
+    done <<< "$ASSET_URLS"
+  fi
+}
+
+function main() {
+  check_vars
+  download_artifacts_from_github
+  upload_artifacts
+  update_version_info
+}
+
+# Usage example:
+#   AWS_ACCESS_KEY_ID=<your_access_key_id> \
+#   AWS_SECRET_ACCESS_KEY=<your_secret_access_key> \
+#   AWS_DEFAULT_REGION=<your_region> \
+#   UPDATE_VERSION_INFO=true \
+#   DOWNLOAD_ARTIFACTS_FROM_GITHUB=false \
+#     ./upload-artifacts-to-s3.sh <artifacts-dir> <version> <aws-s3-bucket>
+main

.github/workflows/apidoc.yml

@@ -17,7 +17,7 @@ env:
 
 jobs:
   apidoc:
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-20.04
     steps:
     - uses: actions/checkout@v3
     - uses: arduino/setup-protoc@v1

.github/workflows/dev-build.yml

@@ -16,11 +16,11 @@ on:
         description: The runner uses to build linux-amd64 artifacts
         default: ec2-c6i.4xlarge-amd64
         options:
-          - ubuntu-latest
-          - ubuntu-latest-8-cores
-          - ubuntu-latest-16-cores
-          - ubuntu-latest-32-cores
-          - ubuntu-latest-64-cores
+          - ubuntu-20.04
+          - ubuntu-20.04-8-cores
+          - ubuntu-20.04-16-cores
+          - ubuntu-20.04-32-cores
+          - ubuntu-20.04-64-cores
           - ec2-c6i.xlarge-amd64 # 4C8G
           - ec2-c6i.2xlarge-amd64 # 8C16G
           - ec2-c6i.4xlarge-amd64 # 16C32G
@@ -78,7 +78,7 @@ jobs:
   allocate-runners:
     name: Allocate runners
     if: ${{ github.repository == 'GreptimeTeam/greptimedb' }}
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-20.04
     outputs:
       linux-amd64-runner: ${{ steps.start-linux-amd64-runner.outputs.label }}
       linux-arm64-runner: ${{ steps.start-linux-arm64-runner.outputs.label }}
@@ -164,12 +164,7 @@ jobs:
           cargo-profile: ${{ env.CARGO_PROFILE }}
           version: ${{ needs.allocate-runners.outputs.version }}
           disable-run-tests: ${{ env.DISABLE_RUN_TESTS }}
-          release-to-s3-bucket: ${{ vars.AWS_RELEASE_BUCKET }}
-          aws-access-key-id: ${{ secrets.AWS_CN_ACCESS_KEY_ID }}
-          aws-secret-access-key: ${{ secrets.AWS_CN_SECRET_ACCESS_KEY }}
-          aws-region: ${{ vars.AWS_RELEASE_BUCKET_REGION }}
           dev-mode: true # Only build the standard greptime binary.
-          upload-to-s3: false # No need to upload to S3.
           working-dir: ${{ env.CHECKOUT_GREPTIMEDB_PATH }}
 
   build-linux-arm64-artifacts:
@@ -198,12 +193,7 @@ jobs:
           cargo-profile: ${{ env.CARGO_PROFILE }}
           version: ${{ needs.allocate-runners.outputs.version }}
           disable-run-tests: ${{ env.DISABLE_RUN_TESTS }}
-          release-to-s3-bucket: ${{ vars.AWS_RELEASE_BUCKET }}
-          aws-access-key-id: ${{ secrets.AWS_CN_ACCESS_KEY_ID }}
-          aws-secret-access-key: ${{ secrets.AWS_CN_SECRET_ACCESS_KEY }}
-          aws-region: ${{ vars.AWS_RELEASE_BUCKET_REGION }}
           dev-mode: true # Only build the standard greptime binary.
-          upload-to-s3: false # No need to upload to S3.
           working-dir: ${{ env.CHECKOUT_GREPTIMEDB_PATH }}
 
   release-images-to-dockerhub:
@@ -214,7 +204,7 @@ jobs:
       build-linux-amd64-artifacts,
       build-linux-arm64-artifacts,
     ]
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-20.04
     outputs:
       build-result: ${{ steps.set-build-result.outputs.build-result }}
     steps:
@@ -239,41 +229,44 @@ jobs:
         run: |
           echo "build-result=success" >> $GITHUB_OUTPUT
 
-  release-images-to-acr:
-    name: Build and push images to ACR
+  release-cn-artifacts:
+    name: Release artifacts to CN region
     if: ${{ inputs.release_images || github.event_name == 'schedule' }}
     needs: [
       allocate-runners,
-      build-linux-amd64-artifacts,
-      build-linux-arm64-artifacts,
+      release-images-to-dockerhub,
     ]
-    runs-on: ubuntu-latest
-    # When we push to ACR, it's easy to fail due to some unknown network issues.
-    # However, we don't want to fail the whole workflow because of this.
-    # The ACR have daily sync with DockerHub, so don't worry about the image not being updated.
+    runs-on: ubuntu-20.04
     continue-on-error: true
     steps:
       - uses: actions/checkout@v3
         with:
           fetch-depth: 0
 
-      - name: Build and push images to ACR
-        uses: ./.github/actions/build-images
+      - name: Release artifacts to CN region
+        uses: ./.github/actions/release-cn-artifacts
         with:
-          image-registry: ${{ vars.ACR_IMAGE_REGISTRY }}
-          image-namespace: ${{ vars.IMAGE_NAMESPACE }}
-          image-name: ${{ env.IMAGE_NAME }}
-          image-registry-username: ${{ secrets.ALICLOUD_USERNAME }}
-          image-registry-password: ${{ secrets.ALICLOUD_PASSWORD }}
+          src-image-registry: docker.io
+          src-image-namespace: ${{ vars.IMAGE_NAMESPACE }}
+          src-image-name: ${{ env.IMAGE_NAME }}
+          dst-image-registry-username: ${{ secrets.ALICLOUD_USERNAME }}
+          dst-image-registry-password: ${{ secrets.ALICLOUD_PASSWORD }}
+          dst-image-registry: ${{ vars.ACR_IMAGE_REGISTRY }}
+          dst-image-namespace: ${{ vars.IMAGE_NAMESPACE }}
           version: ${{ needs.allocate-runners.outputs.version }}
-          push-latest-tag: false # Don't push the latest tag to registry.
-          dev-mode: true # Only build the standard images.
+          aws-cn-s3-bucket: ${{ vars.AWS_RELEASE_BUCKET }}
+          aws-cn-access-key-id: ${{ secrets.AWS_CN_ACCESS_KEY_ID }}
+          aws-cn-secret-access-key: ${{ secrets.AWS_CN_SECRET_ACCESS_KEY }}
+          aws-cn-region: ${{ vars.AWS_RELEASE_BUCKET_REGION }}
+          dev-mode: true                     # Only build the standard images(exclude centos images).
+          push-latest-tag: false             # Don't push the latest tag to registry.
+          update-version-info: false         # Don't update the version info in S3.
 
   stop-linux-amd64-runner: # It's always run as the last job in the workflow to make sure that the runner is released.
     name: Stop linux-amd64 runner
     # Only run this job when the runner is allocated.
     if: ${{ always() }}
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-20.04
     needs: [
       allocate-runners,
       build-linux-amd64-artifacts,
@@ -298,7 +291,7 @@ jobs:
     name: Stop linux-arm64 runner
     # Only run this job when the runner is allocated.
     if: ${{ always() }}
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-20.04
     needs: [
       allocate-runners,
       build-linux-arm64-artifacts,
@@ -325,7 +318,7 @@ jobs:
     needs: [
       release-images-to-dockerhub
     ]
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-20.04
     env:
       SLACK_WEBHOOK_URL: ${{ secrets.SLACK_WEBHOOK_URL_DEVELOP_CHANNEL }}
     steps:

.github/workflows/develop.yml

@@ -34,7 +34,7 @@ env:
 jobs:
   typos:
     name: Spell Check with Typos
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-20.04
     steps:
       - uses: actions/checkout@v3
       - uses: crate-ci/typos@v1.13.10
@@ -42,7 +42,7 @@ jobs:
   check:
     name: Check
     if: github.event.pull_request.draft == false
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-20.04
     timeout-minutes: 60
     steps:
       - uses: actions/checkout@v3
@@ -60,7 +60,7 @@ jobs:
   toml:
     name: Toml Check
     if: github.event.pull_request.draft == false
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-20.04
     timeout-minutes: 60
     steps:
       - uses: actions/checkout@v3
@@ -80,7 +80,7 @@ jobs:
     runs-on: ${{ matrix.os }}
     strategy:
       matrix:
-        os: [ ubuntu-latest-8-cores, windows-latest-8-cores ]
+        os: [ ubuntu-20.04-8-cores ]
     timeout-minutes: 60
     steps:
       - uses: actions/checkout@v3
@@ -105,7 +105,7 @@ jobs:
   fmt:
     name: Rustfmt
     if: github.event.pull_request.draft == false
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-20.04
     timeout-minutes: 60
     steps:
       - uses: actions/checkout@v3
@@ -124,7 +124,7 @@ jobs:
   clippy:
     name: Clippy
     if: github.event.pull_request.draft == false
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-20.04
     timeout-minutes: 60
     steps:
       - uses: actions/checkout@v3
@@ -142,7 +142,7 @@ jobs:
 
   coverage:
     if: github.event.pull_request.draft == false
-    runs-on: ubuntu-latest-8-cores
+    runs-on: ubuntu-20.04-8-cores
     timeout-minutes: 60
     steps:
       - uses: actions/checkout@v3
@@ -188,43 +188,3 @@ jobs:
           flags: rust
           fail_ci_if_error: false
           verbose: true
-
-  test-on-windows:
-    if: github.event.pull_request.draft == false
-    runs-on: windows-latest-8-cores
-    timeout-minutes: 60
-    steps:
-      - run: git config --global core.autocrlf false
-      - uses: actions/checkout@v3
-      - uses: arduino/setup-protoc@v1
-        with:
-          repo-token: ${{ secrets.GITHUB_TOKEN }}
-      - name: Install Rust toolchain
-        uses: dtolnay/rust-toolchain@master
-        with:
-          toolchain: ${{ env.RUST_TOOLCHAIN }}
-          components: llvm-tools-preview
-      - name: Rust Cache
-        uses: Swatinem/rust-cache@v2
-      - name: Install Cargo Nextest
-        uses: taiki-e/install-action@nextest
-      - name: Install Python
-        uses: actions/setup-python@v4
-        with:
-          python-version: '3.10'
-      - name: Install PyArrow Package
-        run: pip install pyarrow
-      - name: Install WSL distribution
-        uses: Vampire/setup-wsl@v2
-        with:
-         distribution: Ubuntu-22.04
-      - name: Running tests
-        run: cargo nextest run -F pyo3_backend,dashboard
-        env:
-          RUST_BACKTRACE: 1
-          CARGO_INCREMENTAL: 0
-          GT_S3_BUCKET: ${{ secrets.S3_BUCKET }}
-          GT_S3_ACCESS_KEY_ID: ${{ secrets.S3_ACCESS_KEY_ID }}
-          GT_S3_ACCESS_KEY: ${{ secrets.S3_ACCESS_KEY }}
-          GT_S3_REGION: ${{ secrets.S3_REGION }}
-          UNITTEST_LOG_DIR: "__unittest_logs"

.github/workflows/doc-issue.yml

@@ -11,7 +11,7 @@ on:
 jobs:
   doc_issue:
     if: github.event.label.name == 'doc update required'
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-20.04
     steps:
       - name: create an issue in doc repo
         uses: dacbd/create-issue-action@main
@@ -25,7 +25,7 @@ jobs:
             ${{ github.event.issue.html_url || github.event.pull_request.html_url }}
   cloud_issue:
     if: github.event.label.name == 'cloud followup required'
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-20.04
     steps:
       - name: create an issue in cloud repo
         uses: dacbd/create-issue-action@main

.github/workflows/docs.yml

@@ -30,7 +30,7 @@ name: CI
 jobs:
   typos:
     name: Spell Check with Typos
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-20.04
     steps:
       - uses: actions/checkout@v3
       - uses: crate-ci/typos@v1.13.10
@@ -38,33 +38,33 @@ jobs:
   check:
     name: Check
     if: github.event.pull_request.draft == false
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-20.04
     steps:
       - run: 'echo "No action required"'
 
   fmt:
     name: Rustfmt
     if: github.event.pull_request.draft == false
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-20.04
     steps:
       - run: 'echo "No action required"'
 
   clippy:
     name: Clippy
     if: github.event.pull_request.draft == false
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-20.04
     steps:
       - run: 'echo "No action required"'
 
   coverage:
     if: github.event.pull_request.draft == false
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-20.04
     steps:
       - run: 'echo "No action required"'
 
   sqlness:
     name: Sqlness Test
     if: github.event.pull_request.draft == false
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-20.04
     steps:
       - run: 'echo "No action required"'

.github/workflows/license.yaml

@@ -8,7 +8,7 @@ on:
     types: [opened, synchronize, reopened, ready_for_review]
 jobs:
   license-header-check:
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-20.04
     name: license-header-check
     steps:
     - uses: actions/checkout@v2

.github/workflows/nightly-build.yml

@@ -14,11 +14,11 @@ on:
         description: The runner uses to build linux-amd64 artifacts
         default: ec2-c6i.2xlarge-amd64
         options:
-          - ubuntu-latest
-          - ubuntu-latest-8-cores
-          - ubuntu-latest-16-cores
-          - ubuntu-latest-32-cores
-          - ubuntu-latest-64-cores
+          - ubuntu-20.04
+          - ubuntu-20.04-8-cores
+          - ubuntu-20.04-16-cores
+          - ubuntu-20.04-32-cores
+          - ubuntu-20.04-64-cores
           - ec2-c6i.xlarge-amd64 # 4C8G
           - ec2-c6i.2xlarge-amd64 # 8C16G
           - ec2-c6i.4xlarge-amd64 # 16C32G
@@ -70,7 +70,7 @@ jobs:
   allocate-runners:
     name: Allocate runners
     if: ${{ github.repository == 'GreptimeTeam/greptimedb' }}
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-20.04
     outputs:
       linux-amd64-runner: ${{ steps.start-linux-amd64-runner.outputs.label }}
       linux-arm64-runner: ${{ steps.start-linux-arm64-runner.outputs.label }}
@@ -147,11 +147,6 @@ jobs:
           cargo-profile: ${{ env.CARGO_PROFILE }}
           version: ${{ needs.allocate-runners.outputs.version }}
           disable-run-tests: ${{ env.DISABLE_RUN_TESTS }}
-          release-to-s3-bucket: ${{ vars.AWS_RELEASE_BUCKET }}
-          aws-access-key-id: ${{ secrets.AWS_CN_ACCESS_KEY_ID }}
-          aws-secret-access-key: ${{ secrets.AWS_CN_SECRET_ACCESS_KEY }}
-          aws-region: ${{ vars.AWS_RELEASE_BUCKET_REGION }}
-          upload-latest-artifacts: false
 
   build-linux-arm64-artifacts:
     name: Build linux-arm64 artifacts
@@ -171,11 +166,6 @@ jobs:
           cargo-profile: ${{ env.CARGO_PROFILE }}
           version: ${{ needs.allocate-runners.outputs.version }}
           disable-run-tests: ${{ env.DISABLE_RUN_TESTS }}
-          release-to-s3-bucket: ${{ vars.AWS_RELEASE_BUCKET }}
-          aws-access-key-id: ${{ secrets.AWS_CN_ACCESS_KEY_ID }}
-          aws-secret-access-key: ${{ secrets.AWS_CN_SECRET_ACCESS_KEY }}
-          aws-region: ${{ vars.AWS_RELEASE_BUCKET_REGION }}
-          upload-latest-artifacts: false
 
   release-images-to-dockerhub:
     name: Build and push images to DockerHub
@@ -185,7 +175,7 @@ jobs:
       build-linux-amd64-artifacts,
       build-linux-arm64-artifacts,
     ]
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-20.04
     outputs:
       nightly-build-result: ${{ steps.set-nightly-build-result.outputs.nightly-build-result }}
     steps:
@@ -208,15 +198,14 @@ jobs:
         run: |
           echo "nightly-build-result=success" >> $GITHUB_OUTPUT
 
-  release-images-to-acr:
-    name: Build and push images to ACR
+  release-cn-artifacts:
+    name: Release artifacts to CN region
     if: ${{ inputs.release_images || github.event_name == 'schedule' }}
     needs: [
       allocate-runners,
-      build-linux-amd64-artifacts,
-      build-linux-arm64-artifacts,
+      release-images-to-dockerhub,
     ]
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-20.04
     # When we push to ACR, it's easy to fail due to some unknown network issues.
     # However, we don't want to fail the whole workflow because of this.
     # The ACR have daily sync with DockerHub, so don't worry about the image not being updated.
@@ -226,21 +215,30 @@ jobs:
         with:
           fetch-depth: 0
 
-      - name: Build and push images to ACR
-        uses: ./.github/actions/build-images
+      - name: Release artifacts to CN region
+        uses: ./.github/actions/release-cn-artifacts
         with:
-          image-registry: ${{ vars.ACR_IMAGE_REGISTRY }}
-          image-namespace: ${{ vars.IMAGE_NAMESPACE }}
-          image-registry-username: ${{ secrets.ALICLOUD_USERNAME }}
-          image-registry-password: ${{ secrets.ALICLOUD_PASSWORD }}
+          src-image-registry: docker.io
+          src-image-namespace: ${{ vars.IMAGE_NAMESPACE }}
+          src-image-name: greptimedb
+          dst-image-registry-username: ${{ secrets.ALICLOUD_USERNAME }}
+          dst-image-registry-password: ${{ secrets.ALICLOUD_PASSWORD }}
+          dst-image-registry: ${{ vars.ACR_IMAGE_REGISTRY }}
+          dst-image-namespace: ${{ vars.IMAGE_NAMESPACE }}
           version: ${{ needs.allocate-runners.outputs.version }}
-          push-latest-tag: false # Don't push the latest tag to registry.
+          aws-cn-s3-bucket: ${{ vars.AWS_RELEASE_BUCKET }}
+          aws-cn-access-key-id: ${{ secrets.AWS_CN_ACCESS_KEY_ID }}
+          aws-cn-secret-access-key: ${{ secrets.AWS_CN_SECRET_ACCESS_KEY }}
+          aws-cn-region: ${{ vars.AWS_RELEASE_BUCKET_REGION }}
+          dev-mode: false
+          update-version-info: false  # Don't update version info in S3.
+          push-latest-tag: false      # Don't push the latest tag to registry.
 
   stop-linux-amd64-runner: # It's always run as the last job in the workflow to make sure that the runner is released.
     name: Stop linux-amd64 runner
     # Only run this job when the runner is allocated.
     if: ${{ always() }}
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-20.04
     needs: [
       allocate-runners,
       build-linux-amd64-artifacts,
@@ -265,7 +263,7 @@ jobs:
     name: Stop linux-arm64 runner
     # Only run this job when the runner is allocated.
     if: ${{ always() }}
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-20.04
     needs: [
       allocate-runners,
       build-linux-arm64-artifacts,
@@ -292,7 +290,7 @@ jobs:
     needs: [
       release-images-to-dockerhub
     ]
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-20.04
     env:
       SLACK_WEBHOOK_URL: ${{ secrets.SLACK_WEBHOOK_URL_DEVELOP_CHANNEL }}
     steps:

.github/workflows/nightly-ci.yml

@@ -0,0 +1,98 @@
+# Nightly CI: runs tests every night for our second tier plaforms (Windows)
+
+on:
+  schedule:
+    - cron: '0 23 * * 1-5'
+  workflow_dispatch:
+
+name: Nightly CI
+
+concurrency:
+  group: ${{ github.workflow }}-${{ github.head_ref || github.run_id }}
+  cancel-in-progress: true
+
+env:
+  RUST_TOOLCHAIN: nightly-2023-08-07
+
+jobs:
+  sqlness:
+    name: Sqlness Test
+    runs-on: ${{ matrix.os }}
+    strategy:
+      matrix:
+        os: [ windows-latest-8-cores ]
+    timeout-minutes: 60
+    steps:
+      - uses: actions/checkout@v4.1.0
+      - uses: arduino/setup-protoc@v1
+        with:
+          repo-token: ${{ secrets.GITHUB_TOKEN }}
+      - uses: dtolnay/rust-toolchain@master
+        with:
+          toolchain: ${{ env.RUST_TOOLCHAIN }}
+      - name: Rust Cache
+        uses: Swatinem/rust-cache@v2
+      - name: Run sqlness
+        run: cargo sqlness
+      - name: Notify slack if failed
+        if: failure()
+        uses: slackapi/slack-github-action@v1.23.0
+        env:
+          SLACK_WEBHOOK_URL: ${{ secrets.SLACK_WEBHOOK_URL_DEVELOP_CHANNEL }}
+        with:
+          payload: |
+            {"text": "Nightly CI failed for sqlness tests"}
+      - name: Upload sqlness logs
+        if: always()
+        uses: actions/upload-artifact@v3
+        with:
+          name: sqlness-logs
+          path: ${{ runner.temp }}/greptime-*.log
+          retention-days: 3
+
+  test-on-windows:
+    runs-on: windows-latest-8-cores
+    timeout-minutes: 60
+    steps:
+      - run: git config --global core.autocrlf false
+      - uses: actions/checkout@v4.1.0
+      - uses: arduino/setup-protoc@v1
+        with:
+          repo-token: ${{ secrets.GITHUB_TOKEN }}
+      - name: Install Rust toolchain
+        uses: dtolnay/rust-toolchain@master
+        with:
+          toolchain: ${{ env.RUST_TOOLCHAIN }}
+          components: llvm-tools-preview
+      - name: Rust Cache
+        uses: Swatinem/rust-cache@v2
+      - name: Install Cargo Nextest
+        uses: taiki-e/install-action@nextest
+      - name: Install Python
+        uses: actions/setup-python@v4
+        with:
+          python-version: '3.10'
+      - name: Install PyArrow Package
+        run: pip install pyarrow
+      - name: Install WSL distribution
+        uses: Vampire/setup-wsl@v2
+        with:
+          distribution: Ubuntu-22.04
+      - name: Running tests
+        run: cargo nextest run -F pyo3_backend,dashboard
+        env:
+          RUST_BACKTRACE: 1
+          CARGO_INCREMENTAL: 0
+          GT_S3_BUCKET: ${{ secrets.S3_BUCKET }}
+          GT_S3_ACCESS_KEY_ID: ${{ secrets.S3_ACCESS_KEY_ID }}
+          GT_S3_ACCESS_KEY: ${{ secrets.S3_ACCESS_KEY }}
+          GT_S3_REGION: ${{ secrets.S3_REGION }}
+          UNITTEST_LOG_DIR: "__unittest_logs"
+      - name: Notify slack if failed
+        if: failure()
+        uses: slackapi/slack-github-action@v1.23.0
+        env:
+          SLACK_WEBHOOK_URL: ${{ secrets.SLACK_WEBHOOK_URL_DEVELOP_CHANNEL }}
+        with:
+          payload: |
+            {"text": "Nightly CI failed for cargo test"}

.github/workflows/pr-title-checker.yml

@@ -10,7 +10,7 @@ on:
 
 jobs:
   check:
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-20.04
     timeout-minutes: 10
     steps:
       - uses: thehanimo/pr-title-checker@v1.3.4
@@ -19,7 +19,7 @@ jobs:
           pass_on_octokit_error: false
           configuration_path: ".github/pr-title-checker-config.json"
   breaking:
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-20.04
     timeout-minutes: 10
     steps:
       - uses: thehanimo/pr-title-checker@v1.3.4

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

@@ -0,0 +1,85 @@
+name: Release dev-builder images
+
+on:
+  workflow_dispatch: # Allows you to run this workflow manually.
+    inputs:
+      version:
+        description: Version of the dev-builder
+        required: false
+        default: latest
+      release_dev_builder_ubuntu_image:
+        type: boolean
+        description: Release dev-builder-ubuntu image
+        required: false
+        default: false
+      release_dev_builder_centos_image:
+        type: boolean
+        description: Release dev-builder-centos image
+        required: false
+        default: false
+      release_dev_builder_android_image:
+        type: boolean
+        description: Release dev-builder-android image
+        required: false
+        default: false
+
+jobs:
+  release-dev-builder-images:
+    name: Release dev builder images
+    if: ${{ inputs.release_dev_builder_ubuntu_image || inputs.release_dev_builder_centos_image || inputs.release_dev_builder_android_image }} # Only manually trigger this job.
+    runs-on: ubuntu-20.04-16-cores
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v3
+        with:
+          fetch-depth: 0
+
+      - name: Build and push dev builder images
+        uses: ./.github/actions/build-dev-builder-images
+        with:
+          version: ${{ inputs.version }}
+          dockerhub-image-registry-username: ${{ secrets.DOCKERHUB_USERNAME }}
+          dockerhub-image-registry-token: ${{ secrets.DOCKERHUB_TOKEN }}
+          build-dev-builder-ubuntu: ${{ inputs.release_dev_builder_ubuntu_image }}
+          build-dev-builder-centos: ${{ inputs.release_dev_builder_centos_image }}
+          build-dev-builder-android: ${{ inputs.release_dev_builder_android_image }}
+
+  release-dev-builder-images-cn: # Note: Be careful issue: https://github.com/containers/skopeo/issues/1874 and we decide to use the latest stable skopeo container.
+    name: Release dev builder images to CN region
+    runs-on: ubuntu-20.04
+    needs: [
+      release-dev-builder-images
+    ]
+    steps:
+      - name: Push dev-builder-ubuntu image
+        shell: bash
+        if: ${{ inputs.release_dev_builder_ubuntu_image }}
+        env:
+          DST_REGISTRY_USERNAME: ${{ secrets.ALICLOUD_USERNAME }}
+          DST_REGISTRY_PASSWORD: ${{ secrets.ALICLOUD_PASSWORD }}
+        run: |
+          docker run quay.io/skopeo/stable:latest copy -a docker://docker.io/${{ vars.IMAGE_NAMESPACE }}/dev-builder-ubuntu:${{ inputs.version }} \
+            --dest-creds "$DST_REGISTRY_USERNAME":"$DST_REGISTRY_PASSWORD" \
+            docker://${{ vars.ACR_IMAGE_REGISTRY }}/${{ vars.IMAGE_NAMESPACE }}/dev-builder-ubuntu:${{ inputs.version }}
+
+      - name: Push dev-builder-centos image
+        shell: bash
+        if: ${{ inputs.release_dev_builder_centos_image }}
+        env:
+          DST_REGISTRY_USERNAME: ${{ secrets.ALICLOUD_USERNAME }}
+          DST_REGISTRY_PASSWORD: ${{ secrets.ALICLOUD_PASSWORD }}
+        run: |
+          docker run quay.io/skopeo/stable:latest copy -a docker://docker.io/${{ vars.IMAGE_NAMESPACE }}/dev-builder-centos:${{ inputs.version }} \
+            --dest-creds "$DST_REGISTRY_USERNAME":"$DST_REGISTRY_PASSWORD" \
+            docker://${{ vars.ACR_IMAGE_REGISTRY }}/${{ vars.IMAGE_NAMESPACE }}/dev-builder-centos:${{ inputs.version }}
+
+      - name: Push dev-builder-android image
+        shell: bash
+        if: ${{ inputs.release_dev_builder_android_image }}
+        env:
+          DST_REGISTRY_USERNAME: ${{ secrets.ALICLOUD_USERNAME }}
+          DST_REGISTRY_PASSWORD: ${{ secrets.ALICLOUD_PASSWORD }}
+        run: |
+          docker run quay.io/skopeo/stable:latest copy -a docker://docker.io/${{ vars.IMAGE_NAMESPACE }}/dev-builder-android:${{ inputs.version }} \
+            --dest-creds "$DST_REGISTRY_USERNAME":"$DST_REGISTRY_PASSWORD" \
+            docker://${{ vars.ACR_IMAGE_REGISTRY }}/${{ vars.IMAGE_NAMESPACE }}/dev-builder-android:${{ inputs.version }}

.github/workflows/release.yml

@@ -18,11 +18,11 @@ on:
         description: The runner uses to build linux-amd64 artifacts
         default: ec2-c6i.4xlarge-amd64
         options:
-          - ubuntu-latest
-          - ubuntu-latest-8-cores
-          - ubuntu-latest-16-cores
-          - ubuntu-latest-32-cores
-          - ubuntu-latest-64-cores
+          - ubuntu-20.04
+          - ubuntu-20.04-8-cores
+          - ubuntu-20.04-16-cores
+          - ubuntu-20.04-32-cores
+          - ubuntu-20.04-64-cores
           - ec2-c6i.xlarge-amd64 # 4C8G
           - ec2-c6i.2xlarge-amd64 # 8C16G
           - ec2-c6i.4xlarge-amd64 # 16C32G
@@ -63,7 +63,12 @@ on:
         description: Build macos artifacts
         required: false
         default: false
-      release_artifacts:
+      build_windows_artifacts:
+        type: boolean
+        description: Build Windows artifacts
+        required: false
+        default: false
+      publish_github_release:
         type: boolean
         description: Create GitHub release and upload artifacts
         required: false
@@ -73,11 +78,6 @@ on:
         description: Build and push images to DockerHub and ACR
         required: false
         default: false
-      release_dev_builder_image:
-        type: boolean
-        description: Release dev-builder image
-        required: false
-        default: false
 
 # Use env variables to control all the release process.
 env:
@@ -91,17 +91,18 @@ 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: v0.4.0
+  NEXT_RELEASE_VERSION: v0.5.0
 
 jobs:
   allocate-runners:
     name: Allocate runners
     if: ${{ github.repository == 'GreptimeTeam/greptimedb' }}
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-20.04
     outputs:
       linux-amd64-runner: ${{ steps.start-linux-amd64-runner.outputs.label }}
       linux-arm64-runner: ${{ steps.start-linux-arm64-runner.outputs.label }}
       macos-runner: ${{ inputs.macos_runner || vars.DEFAULT_MACOS_RUNNER }}
+      windows-runner: windows-latest-8-cores
 
       # The following EC2 resource id will be used for resource releasing.
       linux-amd64-ec2-runner-label: ${{ steps.start-linux-amd64-runner.outputs.label }}
@@ -177,11 +178,6 @@ jobs:
           cargo-profile: ${{ env.CARGO_PROFILE }}
           version: ${{ needs.allocate-runners.outputs.version }}
           disable-run-tests: ${{ env.DISABLE_RUN_TESTS }}
-          release-to-s3-bucket: ${{ vars.AWS_RELEASE_BUCKET }}
-          aws-access-key-id: ${{ secrets.AWS_CN_ACCESS_KEY_ID }}
-          aws-secret-access-key: ${{ secrets.AWS_CN_SECRET_ACCESS_KEY }}
-          aws-region: ${{ vars.AWS_RELEASE_BUCKET_REGION }}
-          upload-to-s3: ${{ vars.UPLOAD_TO_S3 }}
 
   build-linux-arm64-artifacts:
     name: Build linux-arm64 artifacts
@@ -201,11 +197,6 @@ jobs:
           cargo-profile: ${{ env.CARGO_PROFILE }}
           version: ${{ needs.allocate-runners.outputs.version }}
           disable-run-tests: ${{ env.DISABLE_RUN_TESTS }}
-          release-to-s3-bucket: ${{ vars.AWS_RELEASE_BUCKET }}
-          aws-access-key-id: ${{ secrets.AWS_CN_ACCESS_KEY_ID }}
-          aws-secret-access-key: ${{ secrets.AWS_CN_SECRET_ACCESS_KEY }}
-          aws-region: ${{ vars.AWS_RELEASE_BUCKET_REGION }}
-          upload-to-s3: ${{ vars.UPLOAD_TO_S3 }}
 
   build-macos-artifacts:
     name: Build macOS artifacts
@@ -247,12 +238,43 @@ jobs:
           features: ${{ matrix.features }}
           version: ${{ needs.allocate-runners.outputs.version }}
           disable-run-tests: ${{ env.DISABLE_RUN_TESTS }}
-          release-to-s3-bucket: ${{ vars.AWS_RELEASE_BUCKET }}
           artifacts-dir: ${{ matrix.artifacts-dir-prefix }}-${{ needs.allocate-runners.outputs.version }}
-          aws-access-key-id: ${{ secrets.AWS_CN_ACCESS_KEY_ID }}
-          aws-secret-access-key: ${{ secrets.AWS_CN_SECRET_ACCESS_KEY }}
-          aws-region: ${{ vars.AWS_RELEASE_BUCKET_REGION }}
-          upload-to-s3: ${{ vars.UPLOAD_TO_S3 }}
+
+  build-windows-artifacts:
+    name: Build Windows artifacts
+    strategy:
+      fail-fast: false
+      matrix:
+        include:
+          - os: ${{ needs.allocate-runners.outputs.windows-runner }}
+            arch: x86_64-pc-windows-msvc
+            features: servers/dashboard
+            artifacts-dir-prefix: greptime-windows-amd64
+          - os: ${{ needs.allocate-runners.outputs.windows-runner }}
+            arch: x86_64-pc-windows-msvc
+            features: pyo3_backend,servers/dashboard
+            artifacts-dir-prefix: greptime-windows-amd64-pyo3
+    runs-on: ${{ matrix.os }}
+    needs: [
+      allocate-runners,
+    ]
+    if: ${{ inputs.build_windows_artifacts || github.event_name == 'push' || github.event_name == 'schedule' }}
+    steps:
+      - run: git config --global core.autocrlf false
+
+      - uses: actions/checkout@v3
+        with:
+          fetch-depth: 0
+
+      - uses: ./.github/actions/build-windows-artifacts
+        with:
+          arch: ${{ matrix.arch }}
+          rust-toolchain: ${{ env.RUST_TOOLCHAIN }}
+          cargo-profile: ${{ env.CARGO_PROFILE }}
+          features: ${{ matrix.features }}
+          version: ${{ needs.allocate-runners.outputs.version }}
+          disable-run-tests: ${{ env.DISABLE_RUN_TESTS }}
+          artifacts-dir: ${{ matrix.artifacts-dir-prefix }}-${{ needs.allocate-runners.outputs.version }}
 
   release-images-to-dockerhub:
     name: Build and push images to DockerHub
@@ -277,15 +299,18 @@ jobs:
           image-registry-password: ${{ secrets.DOCKERHUB_TOKEN }}
           version: ${{ needs.allocate-runners.outputs.version }}
 
-  release-images-to-acr:
-    name: Build and push images to ACR
+  release-cn-artifacts:
+    name: Release artifacts to CN region
     if: ${{ inputs.release_images || github.event_name == 'push' || github.event_name == 'schedule' }}
-    needs: [
+    needs: [ # The job have to wait for all the artifacts are built.
       allocate-runners,
       build-linux-amd64-artifacts,
       build-linux-arm64-artifacts,
+      build-macos-artifacts,
+      build-windows-artifacts,
+      release-images-to-dockerhub,
     ]
-    runs-on: ubuntu-2004-16-cores
+    runs-on: ubuntu-20.04
     # When we push to ACR, it's easy to fail due to some unknown network issues.
     # However, we don't want to fail the whole workflow because of this.
     # The ACR have daily sync with DockerHub, so don't worry about the image not being updated.
@@ -295,55 +320,47 @@ jobs:
         with:
           fetch-depth: 0
 
-      - name: Build and push images to ACR
-        uses: ./.github/actions/build-images
+      - name: Release artifacts to CN region
+        uses: ./.github/actions/release-cn-artifacts
         with:
-          image-registry: ${{ vars.ACR_IMAGE_REGISTRY }}
-          image-namespace: ${{ vars.IMAGE_NAMESPACE }}
-          image-registry-username: ${{ secrets.ALICLOUD_USERNAME }}
-          image-registry-password: ${{ secrets.ALICLOUD_PASSWORD }}
+          src-image-registry: docker.io
+          src-image-namespace: ${{ vars.IMAGE_NAMESPACE }}
+          src-image-name: greptimedb
+          dst-image-registry-username: ${{ secrets.ALICLOUD_USERNAME }}
+          dst-image-registry-password: ${{ secrets.ALICLOUD_PASSWORD }}
+          dst-image-registry: ${{ vars.ACR_IMAGE_REGISTRY }}
+          dst-image-namespace: ${{ vars.IMAGE_NAMESPACE }}
           version: ${{ needs.allocate-runners.outputs.version }}
+          aws-cn-s3-bucket: ${{ vars.AWS_RELEASE_BUCKET }}
+          aws-cn-access-key-id: ${{ secrets.AWS_CN_ACCESS_KEY_ID }}
+          aws-cn-secret-access-key: ${{ secrets.AWS_CN_SECRET_ACCESS_KEY }}
+          aws-cn-region: ${{ vars.AWS_RELEASE_BUCKET_REGION }}
+          dev-mode: false
+          update-version-info: true
+          push-latest-tag: true
 
-  release-artifacts:
+  publish-github-release:
     name: Create GitHub release and upload artifacts
-    if: ${{ inputs.release_artifacts || github.event_name == 'push' || github.event_name == 'schedule' }}
-    needs: [
+    if: ${{ inputs.publish_github_release || github.event_name == 'push' || github.event_name == 'schedule' }}
+    needs: [ # The job have to wait for all the artifacts are built.
       allocate-runners,
       build-linux-amd64-artifacts,
       build-linux-arm64-artifacts,
       build-macos-artifacts,
+      build-windows-artifacts,
       release-images-to-dockerhub,
     ]
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-20.04
     steps:
       - uses: actions/checkout@v3
         with:
           fetch-depth: 0
 
-      - name: Release artifacts
-        uses: ./.github/actions/release-artifacts
+      - name: Publish GitHub release
+        uses: ./.github/actions/publish-github-release
         with:
           version: ${{ needs.allocate-runners.outputs.version }}
 
-  release-dev-builder-image:
-    name: Release dev builder image
-    if: ${{ inputs.release_dev_builder_image }} # Only manually trigger this job.
-    runs-on: ubuntu-latest-16-cores
-    steps:
-      - name: Checkout
-        uses: actions/checkout@v3
-        with:
-          fetch-depth: 0
-
-      - name: Build and push dev builder image
-        uses: ./.github/actions/build-dev-builder-image
-        with:
-          dockerhub-image-registry-username: ${{ secrets.DOCKERHUB_USERNAME }}
-          dockerhub-image-registry-token: ${{ secrets.DOCKERHUB_TOKEN }}
-          acr-image-registry: ${{ vars.ACR_IMAGE_REGISTRY }}
-          acr-image-registry-username: ${{ secrets.ALICLOUD_USERNAME }}
-          acr-image-registry-password: ${{ secrets.ALICLOUD_PASSWORD }}
-
   ### Stop runners ###
   # It's very necessary to split the job of releasing runners into 'stop-linux-amd64-runner' and 'stop-linux-arm64-runner'.
   # Because we can terminate the specified EC2 instance immediately after the job is finished without uncessary waiting.
@@ -351,7 +368,7 @@ jobs:
     name: Stop linux-amd64 runner
     # Only run this job when the runner is allocated.
     if: ${{ always() }}
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-20.04
     needs: [
       allocate-runners,
       build-linux-amd64-artifacts,
@@ -376,7 +393,7 @@ jobs:
     name: Stop linux-arm64 runner
     # Only run this job when the runner is allocated.
     if: ${{ always() }}
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-20.04
     needs: [
       allocate-runners,
       build-linux-arm64-artifacts,

.github/workflows/size-label.yml

@@ -0,0 +1,26 @@
+name: size-labeler
+
+on: [pull_request]
+
+jobs:
+  labeler:
+    runs-on: ubuntu-latest
+    name: Label the PR size
+    steps:
+      - uses: codelytv/pr-size-labeler@v1
+        with:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+          s_label: 'Size: S'
+          s_max_size: '100'
+          m_label: 'Size: M'
+          m_max_size: '500'
+          l_label: 'Size: L'
+          l_max_size: '1000'
+          xl_label: 'Size: XL'
+          fail_if_xl: 'false'
+          message_if_xl: >
+            This PR exceeds the recommended size of 1000 lines.
+            Please make sure you are NOT addressing multiple issues with one PR.
+            Note this PR might be rejected due to its size.
+          github_api_url: 'api.github.com'
+          files_to_ignore: 'Cargo.lock'

Cargo.toml

@@ -39,6 +39,7 @@ members = [
     "src/object-store",
     "src/operator",
     "src/partition",
+    "src/plugins",
     "src/promql",
     "src/query",
     "src/script",
@@ -54,39 +55,43 @@ members = [
 resolver = "2"
 
 [workspace.package]
-version = "0.4.0-nightly"
+version = "0.4.2"
 edition = "2021"
 license = "Apache-2.0"
 
 [workspace.dependencies]
 aquamarine = "0.3"
 arrow = { version = "43.0" }
-etcd-client = "0.11"
 arrow-array = "43.0"
 arrow-flight = "43.0"
 arrow-schema = { version = "43.0", features = ["serde"] }
 async-stream = "0.3"
 async-trait = "0.1"
 chrono = { version = "0.4", features = ["serde"] }
-datafusion = { git = "https://github.com/waynexia/arrow-datafusion.git", rev = "c0b0fca548e99d020c76e1a1cd7132aab26000e1" }
-datafusion-common = { git = "https://github.com/waynexia/arrow-datafusion.git", rev = "c0b0fca548e99d020c76e1a1cd7132aab26000e1" }
-datafusion-expr = { git = "https://github.com/waynexia/arrow-datafusion.git", rev = "c0b0fca548e99d020c76e1a1cd7132aab26000e1" }
-datafusion-optimizer = { git = "https://github.com/waynexia/arrow-datafusion.git", rev = "c0b0fca548e99d020c76e1a1cd7132aab26000e1" }
-datafusion-physical-expr = { git = "https://github.com/waynexia/arrow-datafusion.git", rev = "c0b0fca548e99d020c76e1a1cd7132aab26000e1" }
-datafusion-sql = { git = "https://github.com/waynexia/arrow-datafusion.git", rev = "c0b0fca548e99d020c76e1a1cd7132aab26000e1" }
-datafusion-substrait = { git = "https://github.com/waynexia/arrow-datafusion.git", rev = "c0b0fca548e99d020c76e1a1cd7132aab26000e1" }
+datafusion = { git = "https://github.com/waynexia/arrow-datafusion.git", rev = "b6f3b28b6fe91924cc8dd3d83726b766f2a706ec" }
+datafusion-common = { git = "https://github.com/waynexia/arrow-datafusion.git", rev = "b6f3b28b6fe91924cc8dd3d83726b766f2a706ec" }
+datafusion-expr = { git = "https://github.com/waynexia/arrow-datafusion.git", rev = "b6f3b28b6fe91924cc8dd3d83726b766f2a706ec" }
+datafusion-optimizer = { git = "https://github.com/waynexia/arrow-datafusion.git", rev = "b6f3b28b6fe91924cc8dd3d83726b766f2a706ec" }
+datafusion-physical-expr = { git = "https://github.com/waynexia/arrow-datafusion.git", rev = "b6f3b28b6fe91924cc8dd3d83726b766f2a706ec" }
+datafusion-sql = { git = "https://github.com/waynexia/arrow-datafusion.git", rev = "b6f3b28b6fe91924cc8dd3d83726b766f2a706ec" }
+datafusion-substrait = { git = "https://github.com/waynexia/arrow-datafusion.git", rev = "b6f3b28b6fe91924cc8dd3d83726b766f2a706ec" }
 derive_builder = "0.12"
+etcd-client = "0.11"
 futures = "0.3"
 futures-util = "0.3"
-greptime-proto = { git = "https://github.com/GreptimeTeam/greptime-proto.git", rev = "115c1080773be8a819e50b257fece9f839a0c836" }
+greptime-proto = { git = "https://github.com/GreptimeTeam/greptime-proto.git", rev = "1f1dd532a111e3834cc3019c5605e2993ffb9dc3" }
 humantime-serde = "1.1"
 itertools = "0.10"
 lazy_static = "1.4"
+meter-core = { git = "https://github.com/GreptimeTeam/greptime-meter.git", rev = "abbd357c1e193cd270ea65ee7652334a150b628f" }
+metrics = "0.20"
+moka = "0.12"
 once_cell = "1.18"
-opentelemetry-proto = { version = "0.2", features = ["gen-tonic", "metrics"] }
+opentelemetry-proto = { version = "0.2", features = ["gen-tonic", "metrics", "traces"] }
 parquet = "43.0"
 paste = "1.0"
 prost = "0.11"
+raft-engine = { git = "https://github.com/tikv/raft-engine.git", rev = "22dfb426cd994602b57725ef080287d3e53db479" }
 rand = "0.8"
 regex = "1.8"
 reqwest = { version = "0.11", default-features = false, features = [
@@ -98,7 +103,7 @@ serde = { version = "1.0", features = ["derive"] }
 serde_json = "1.0"
 smallvec = "1"
 snafu = { version = "0.7", features = ["backtraces"] }
-sqlparser = { git = "https://github.com/GreptimeTeam/sqlparser-rs.git", rev = "296a4f6c73b129d6f565a42a2e5e53c6bc2b9da4", features = [
+sqlparser = { git = "https://github.com/GreptimeTeam/sqlparser-rs.git", rev = "6cf9d23d5b8fbecd65efc1d9afb7e80ad7a424da", features = [
     "visitor",
 ] }
 strum = { version = "0.25", features = ["derive"] }
@@ -108,8 +113,6 @@ tokio-util = { version = "0.7", features = ["io-util", "compat"] }
 toml = "0.7"
 tonic = { version = "0.9", features = ["tls"] }
 uuid = { version = "1", features = ["serde", "v4", "fast-rng"] }
-metrics = "0.20"
-meter-core = { git = "https://github.com/GreptimeTeam/greptime-meter.git", rev = "abbd357c1e193cd270ea65ee7652334a150b628f" }
 ## workspaces members
 api = { path = "src/api" }
 auth = { path = "src/auth" }
@@ -122,19 +125,18 @@ common-config = { path = "src/common/config" }
 common-datasource = { path = "src/common/datasource" }
 common-error = { path = "src/common/error" }
 common-function = { path = "src/common/function" }
-common-macro = { path = "src/common/macro" }
 common-greptimedb-telemetry = { path = "src/common/greptimedb-telemetry" }
 common-grpc = { path = "src/common/grpc" }
 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-pprof = { path = "src/common/pprof" }
 common-procedure = { path = "src/common/procedure" }
 common-procedure-test = { path = "src/common/procedure-test" }
-common-pprof = { path = "src/common/pprof" }
 common-query = { path = "src/common/query" }
 common-recordbatch = { path = "src/common/recordbatch" }
 common-runtime = { path = "src/common/runtime" }
-substrait = { path = "src/common/substrait" }
 common-telemetry = { path = "src/common/telemetry" }
 common-test-util = { path = "src/common/test-util" }
 common-time = { path = "src/common/time" }
@@ -148,20 +150,20 @@ meta-client = { path = "src/meta-client" }
 meta-srv = { path = "src/meta-srv" }
 mito = { path = "src/mito" }
 mito2 = { path = "src/mito2" }
-operator = { path = "src/operator" }
 object-store = { path = "src/object-store" }
+operator = { path = "src/operator" }
 partition = { path = "src/partition" }
+plugins = { path = "src/plugins" }
 promql = { path = "src/promql" }
 query = { path = "src/query" }
-raft-engine = { git = "https://github.com/tikv/raft-engine.git", rev = "22dfb426cd994602b57725ef080287d3e53db479" }
 script = { path = "src/script" }
 servers = { path = "src/servers" }
 session = { path = "src/session" }
 sql = { path = "src/sql" }
 storage = { path = "src/storage" }
 store-api = { path = "src/store-api" }
+substrait = { path = "src/common/substrait" }
 table = { path = "src/table" }
-table-procedure = { path = "src/table-procedure" }
 
 [workspace.dependencies.meter-macros]
 git = "https://github.com/GreptimeTeam/greptime-meter.git"

Makefile

@@ -55,11 +55,15 @@ else
 	BUILDX_MULTI_PLATFORM_BUILD_OPTS := -o type=docker
 endif
 
+ifneq ($(strip $(CARGO_BUILD_EXTRA_OPTS)),)
+	CARGO_BUILD_OPTS += ${CARGO_BUILD_EXTRA_OPTS}
+endif
+
 ##@ Build
 
 .PHONY: build
 build: ## Build debug version greptime.
-	cargo build ${CARGO_BUILD_OPTS}
+	cargo ${CARGO_EXTENSION} build ${CARGO_BUILD_OPTS}
 
 .POHNY: build-by-dev-builder
 build-by-dev-builder: ## Build greptime by dev-builder.
@@ -67,11 +71,34 @@ build-by-dev-builder: ## Build greptime by dev-builder.
 	-v ${PWD}:/greptimedb -v ${CARGO_REGISTRY_CACHE}:/root/.cargo/registry \
 	-w /greptimedb ${IMAGE_REGISTRY}/${IMAGE_NAMESPACE}/dev-builder-${BASE_IMAGE}:latest \
 	make build \
+	CARGO_EXTENSION="${CARGO_EXTENSION}" \
 	CARGO_PROFILE=${CARGO_PROFILE} \
 	FEATURES=${FEATURES} \
 	TARGET_DIR=${TARGET_DIR} \
 	TARGET=${TARGET} \
-	RELEASE=${RELEASE}
+	RELEASE=${RELEASE} \
+	CARGO_BUILD_EXTRA_OPTS="${CARGO_BUILD_EXTRA_OPTS}"
+
+.PHONY: build-android-bin
+build-android-bin: ## Build greptime binary for android.
+	docker run --network=host \
+	-v ${PWD}:/greptimedb -v ${CARGO_REGISTRY_CACHE}:/root/.cargo/registry \
+	-w /greptimedb ${IMAGE_REGISTRY}/${IMAGE_NAMESPACE}/dev-builder-android:latest \
+	make build \
+	CARGO_EXTENSION="ndk --platform 23 -t aarch64-linux-android" \
+	CARGO_PROFILE=release \
+	FEATURES="${FEATURES}" \
+	TARGET_DIR="${TARGET_DIR}" \
+	TARGET="${TARGET}" \
+	RELEASE="${RELEASE}" \
+	CARGO_BUILD_EXTRA_OPTS="--bin greptime --no-default-features"
+
+.PHONY: strip-android-bin
+strip-android-bin: build-android-bin ## Strip greptime binary for android.
+	docker run --network=host \
+	-v ${PWD}:/greptimedb \
+	-w /greptimedb ${IMAGE_REGISTRY}/${IMAGE_NAMESPACE}/dev-builder-android:latest \
+	bash -c '$${NDK_ROOT}/toolchains/llvm/prebuilt/linux-x86_64/bin/llvm-strip /greptimedb/target/aarch64-linux-android/release/greptime'
 
 .PHONY: clean
 clean: ## Clean the project.

README.md

@@ -27,6 +27,14 @@
     <a href="https://greptime.com/slack"><img src="https://img.shields.io/badge/slack-GreptimeDB-0abd59?logo=slack" alt="slack" /></a>
 </p>
 
+## Upcoming Event
+Come and meet us in **KubeCon + CloudNativeCon North America 2023!**
+<p align="center">
+  <picture>
+    <img alt="KubeCon + CloudNativeCon North Logo" src="./docs/banner/KCCNC_NA_2023_1000x200_Email Banner.png" width="800px">
+  </picture>
+</p>
+
 ## What is GreptimeDB
 
 GreptimeDB is an open-source time-series database with a special focus on
@@ -96,11 +104,11 @@ Or if you built from docker:
 docker run -p 4002:4002 -v "$(pwd):/tmp/greptimedb" greptime/greptimedb standalone start
 ```
 
-Please see [the online document site](https://docs.greptime.com/getting-started/overview#install-greptimedb) for more installation options and [operations info](https://docs.greptime.com/user-guide/operations/overview).
+Please see the online document site for more installation options and [operations info](https://docs.greptime.com/user-guide/operations/overview).
 
 ### Get started
 
-Read the [complete getting started guide](https://docs.greptime.com/getting-started/overview#connect) on our [official document site](https://docs.greptime.com/).
+Read the [complete getting started guide](https://docs.greptime.com/getting-started/try-out-greptimedb) on our [official document site](https://docs.greptime.com/).
 
 To write and query data, GreptimeDB is compatible with multiple [protocols and clients](https://docs.greptime.com/user-guide/clients/overview).
 

benchmarks/Cargo.toml

@@ -6,8 +6,10 @@ license.workspace = true
 
 [dependencies]
 arrow.workspace = true
+chrono.workspace = true
 clap = { version = "4.0", features = ["derive"] }
 client = { workspace = true }
+futures-util.workspace = true
 indicatif = "0.17.1"
 itertools.workspace = true
 parquet.workspace = true

benchmarks/src/bin/nyc-taxi.rs

@@ -29,14 +29,14 @@ use client::api::v1::column::Values;
 use client::api::v1::{
     Column, ColumnDataType, ColumnDef, CreateTableExpr, InsertRequest, InsertRequests, SemanticType,
 };
-use client::{Client, Database, DEFAULT_CATALOG_NAME, DEFAULT_SCHEMA_NAME};
+use client::{Client, Database, Output, DEFAULT_CATALOG_NAME, DEFAULT_SCHEMA_NAME};
+use futures_util::TryStreamExt;
 use indicatif::{MultiProgress, ProgressBar, ProgressStyle};
 use parquet::arrow::arrow_reader::ParquetRecordBatchReaderBuilder;
 use tokio::task::JoinSet;
 
 const CATALOG_NAME: &str = "greptime";
 const SCHEMA_NAME: &str = "public";
-const TABLE_NAME: &str = "nyc_taxi";
 
 #[derive(Parser)]
 #[command(name = "NYC benchmark runner")]
@@ -74,7 +74,12 @@ fn get_file_list<P: AsRef<Path>>(path: P) -> Vec<PathBuf> {
         .collect()
 }
 
+fn new_table_name() -> String {
+    format!("nyc_taxi_{}", chrono::Utc::now().timestamp())
+}
+
 async fn write_data(
+    table_name: &str,
     batch_size: usize,
     db: &Database,
     path: PathBuf,
@@ -104,7 +109,7 @@ async fn write_data(
         }
         let (columns, row_count) = convert_record_batch(record_batch);
         let request = InsertRequest {
-            table_name: TABLE_NAME.to_string(),
+            table_name: table_name.to_string(),
             columns,
             row_count,
         };
@@ -113,7 +118,7 @@ async fn write_data(
         };
 
         let now = Instant::now();
-        let _ = db.insert(requests).await.unwrap();
+        db.insert(requests).await.unwrap();
         let elapsed = now.elapsed();
         total_rpc_elapsed_ms += elapsed.as_millis();
         progress_bar.inc(row_count as _);
@@ -131,6 +136,11 @@ fn convert_record_batch(record_batch: RecordBatch) -> (Vec<Column>, u32) {
 
     for (array, field) in record_batch.columns().iter().zip(fields.iter()) {
         let (values, datatype) = build_values(array);
+        let semantic_type = match field.name().as_str() {
+            "VendorID" => SemanticType::Tag,
+            "tpep_pickup_datetime" => SemanticType::Timestamp,
+            _ => SemanticType::Field,
+        };
 
         let column = Column {
             column_name: field.name().clone(),
@@ -141,8 +151,7 @@ fn convert_record_batch(record_batch: RecordBatch) -> (Vec<Column>, u32) {
                 .map(|bitmap| bitmap.buffer().as_slice().to_vec())
                 .unwrap_or_default(),
             datatype: datatype.into(),
-            // datatype and semantic_type are set to default
-            ..Default::default()
+            semantic_type: semantic_type as i32,
         };
         columns.push(column);
     }
@@ -243,11 +252,11 @@ fn is_record_batch_full(batch: &RecordBatch) -> bool {
     batch.columns().iter().all(|col| col.null_count() == 0)
 }
 
-fn create_table_expr() -> CreateTableExpr {
+fn create_table_expr(table_name: &str) -> CreateTableExpr {
     CreateTableExpr {
         catalog_name: CATALOG_NAME.to_string(),
         schema_name: SCHEMA_NAME.to_string(),
-        table_name: TABLE_NAME.to_string(),
+        table_name: table_name.to_string(),
         desc: "".to_string(),
         column_defs: vec![
             ColumnDef {
@@ -261,7 +270,7 @@ fn create_table_expr() -> CreateTableExpr {
             ColumnDef {
                 name: "tpep_pickup_datetime".to_string(),
                 data_type: ColumnDataType::TimestampMicrosecond as i32,
-                is_nullable: true,
+                is_nullable: false,
                 default_constraint: vec![],
                 semantic_type: SemanticType::Timestamp as i32,
                 comment: String::new(),
@@ -405,31 +414,31 @@ fn create_table_expr() -> CreateTableExpr {
         ],
         time_index: "tpep_pickup_datetime".to_string(),
         primary_keys: vec!["VendorID".to_string()],
-        create_if_not_exists: false,
+        create_if_not_exists: true,
         table_options: Default::default(),
         table_id: None,
         engine: "mito".to_string(),
     }
 }
 
-fn query_set() -> HashMap<String, String> {
+fn query_set(table_name: &str) -> HashMap<String, String> {
     HashMap::from([
         (
             "count_all".to_string(), 
-            format!("SELECT COUNT(*) FROM {TABLE_NAME};"),
+            format!("SELECT COUNT(*) FROM {table_name};"),
         ),
         (
             "fare_amt_by_passenger".to_string(),
-            format!("SELECT passenger_count, MIN(fare_amount), MAX(fare_amount), SUM(fare_amount) FROM {TABLE_NAME} GROUP BY passenger_count"),
+            format!("SELECT passenger_count, MIN(fare_amount), MAX(fare_amount), SUM(fare_amount) FROM {table_name} GROUP BY passenger_count"),
         )
     ])
 }
 
-async fn do_write(args: &Args, db: &Database) {
+async fn do_write(args: &Args, db: &Database, table_name: &str) {
     let mut file_list = get_file_list(args.path.clone().expect("Specify data path in argument"));
     let mut write_jobs = JoinSet::new();
 
-    let create_table_result = db.create(create_table_expr()).await;
+    let create_table_result = db.create(create_table_expr(table_name)).await;
     println!("Create table result: {create_table_result:?}");
 
     let progress_bar_style = ProgressStyle::with_template(
@@ -447,8 +456,10 @@ async fn do_write(args: &Args, db: &Database) {
             let db = db.clone();
             let mpb = multi_progress_bar.clone();
             let pb_style = progress_bar_style.clone();
-            let _ = write_jobs
-                .spawn(async move { write_data(batch_size, &db, path, mpb, pb_style).await });
+            let table_name = table_name.to_string();
+            let _ = write_jobs.spawn(async move {
+                write_data(&table_name, batch_size, &db, path, mpb, pb_style).await
+            });
         }
     }
     while write_jobs.join_next().await.is_some() {
@@ -457,24 +468,32 @@ async fn do_write(args: &Args, db: &Database) {
             let db = db.clone();
             let mpb = multi_progress_bar.clone();
             let pb_style = progress_bar_style.clone();
-            let _ = write_jobs
-                .spawn(async move { write_data(batch_size, &db, path, mpb, pb_style).await });
+            let table_name = table_name.to_string();
+            let _ = write_jobs.spawn(async move {
+                write_data(&table_name, batch_size, &db, path, mpb, pb_style).await
+            });
         }
     }
 }
 
-async fn do_query(num_iter: usize, db: &Database) {
-    for (query_name, query) in query_set() {
+async fn do_query(num_iter: usize, db: &Database, table_name: &str) {
+    for (query_name, query) in query_set(table_name) {
         println!("Running query: {query}");
         for i in 0..num_iter {
             let now = Instant::now();
-            let _res = db.sql(&query).await.unwrap();
+            let res = db.sql(&query).await.unwrap();
+            match res {
+                Output::AffectedRows(_) | Output::RecordBatches(_) => (),
+                Output::Stream(stream) => {
+                    stream.try_collect::<Vec<_>>().await.unwrap();
+                }
+            }
             let elapsed = now.elapsed();
             println!(
                 "query {}, iteration {}: {}ms",
                 query_name,
                 i,
-                elapsed.as_millis()
+                elapsed.as_millis(),
             );
         }
     }
@@ -491,13 +510,14 @@ fn main() {
         .block_on(async {
             let client = Client::with_urls(vec![&args.endpoint]);
             let db = Database::new(DEFAULT_CATALOG_NAME, DEFAULT_SCHEMA_NAME, client);
+            let table_name = new_table_name();
 
             if !args.skip_write {
-                do_write(&args, &db).await;
+                do_write(&args, &db, &table_name).await;
             }
 
             if !args.skip_read {
-                do_query(args.iter_num, &db).await;
+                do_query(args.iter_num, &db, &table_name).await;
             }
         })
 }

config/datanode.example.toml

@@ -13,17 +13,19 @@ rpc_runtime_size = 8
 require_lease_before_startup = false
 
 [heartbeat]
-# Interval for sending heartbeat messages to the Metasrv in milliseconds, 5000 by default.
-interval_millis = 5000
+# Interval for sending heartbeat messages to the Metasrv, 3 seconds by default.
+interval = "3s"
 
 # Metasrv client options.
 [meta_client]
 # Metasrv address list.
 metasrv_addrs = ["127.0.0.1:3002"]
-# Operation timeout in milliseconds, 3000 by default.
-timeout_millis = 3000
-# Connect server timeout in milliseconds, 5000 by default.
-connect_timeout_millis = 5000
+# Heartbeat timeout, 500 milliseconds by default.
+heartbeat_timeout = "500ms"
+# Operation timeout, 3 seconds by default.
+timeout = "3s"
+# Connect server timeout, 1 second by default.
+connect_timeout = "1s"
 # `TCP_NODELAY` option for accepted connections, true by default.
 tcp_nodelay = true
 
@@ -45,6 +47,12 @@ type = "File"
 # TTL for all tables. Disabled by default.
 # global_ttl = "7d"
 
+# Cache configuration for object storage such as 'S3' etc.
+# The local file cache directory
+# cache_path = "/path/local_cache"
+# The local file cache capacity in bytes.
+# cache_capacity = "256MB"
+
 # Compaction options, see `standalone.example.toml`.
 [storage.compaction]
 max_inflight_tasks = 4

config/frontend.example.toml

@@ -2,10 +2,10 @@
 mode = "distributed"
 
 [heartbeat]
-# Interval for sending heartbeat task to the Metasrv in milliseconds, 5000 by default.
-interval_millis = 5000
-# Interval for retry sending heartbeat task in milliseconds, 5000 by default.
-retry_interval_millis = 5000
+# Interval for sending heartbeat task to the Metasrv, 5 seconds by default.
+interval = "5s"
+# Interval for retry sending heartbeat task, 5 seconds by default.
+retry_interval = "5s"
 
 # HTTP server options, see `standalone.example.toml`.
 [http]
@@ -59,10 +59,10 @@ enable = true
 # Metasrv client options, see `datanode.example.toml`.
 [meta_client]
 metasrv_addrs = ["127.0.0.1:3002"]
-timeout_millis = 3000
+timeout = "3s"
 # DDL timeouts options.
-ddl_timeout_millis = 10000
-connect_timeout_millis = 5000
+ddl_timeout = "10s"
+connect_timeout = "1s"
 tcp_nodelay = true
 
 # Log options, see `standalone.example.toml`

config/metasrv.example.toml

@@ -32,6 +32,6 @@ retry_delay = "500ms"
 # [datanode]
 # # Datanode client options.
 # [datanode.client_options]
-# timeout_millis = 10000
-# connect_timeout_millis = 10000
+# timeout = "10s"
+# connect_timeout = "10s"
 # tcp_nodelay = true

config/standalone.example.toml

@@ -82,6 +82,8 @@ enable = true
 
 # WAL options.
 [wal]
+# WAL data directory
+# dir = "/tmp/greptimedb/wal"
 # WAL file size in bytes.
 file_size = "256MB"
 # WAL purge threshold.
@@ -93,8 +95,8 @@ read_batch_size = 128
 # Whether to sync log file after every write.
 sync_write = false
 
-# Kv options.
-[kv_store]
+# Metadata storage options.
+[metadata_store]
 # Kv file size in bytes.
 file_size = "256MB"
 # Kv purge threshold.
@@ -115,6 +117,10 @@ data_home = "/tmp/greptimedb/"
 type = "File"
 # TTL for all tables. Disabled by default.
 # global_ttl = "7d"
+# Cache configuration for object storage such as 'S3' etc.
+# cache_path = "/path/local_cache"
+# The local file cache capacity in bytes.
+# cache_capacity = "256MB"
 
 # Compaction options.
 [storage.compaction]

docker/buildx/ubuntu/Dockerfile

@@ -1,4 +1,4 @@
-FROM ubuntu:22.04 as builder
+FROM ubuntu:20.04 as builder
 
 ARG CARGO_PROFILE
 ARG FEATURES
@@ -7,6 +7,11 @@ ARG OUTPUT_DIR
 ENV LANG en_US.utf8
 WORKDIR /greptimedb
 
+# Add PPA for Python 3.10.
+RUN apt-get update && \
+    DEBIAN_FRONTEND=noninteractive apt-get install -y software-properties-common && \
+    add-apt-repository ppa:deadsnakes/ppa -y
+
 # Install dependencies.
 RUN --mount=type=cache,target=/var/cache/apt \
     apt-get update && apt-get install -y \

docker/dev-builder/android/Dockerfile

@@ -0,0 +1,41 @@
+FROM --platform=linux/amd64 saschpe/android-ndk:34-jdk17.0.8_7-ndk25.2.9519653-cmake3.22.1
+
+ENV LANG en_US.utf8
+WORKDIR /greptimedb
+
+# Rename libunwind to libgcc
+RUN cp ${NDK_ROOT}/toolchains/llvm/prebuilt/linux-x86_64/lib64/clang/14.0.7/lib/linux/aarch64/libunwind.a ${NDK_ROOT}/toolchains/llvm/prebuilt/linux-x86_64/lib64/clang/14.0.7/lib/linux/aarch64/libgcc.a
+
+# Install dependencies.
+RUN apt-get update && apt-get install -y \
+    libssl-dev \
+    protobuf-compiler \
+    curl \
+    git \
+    build-essential \
+    pkg-config \
+    python3 \
+    python3-dev \
+    python3-pip \
+    && pip3 install --upgrade pip \
+    && pip3 install pyarrow
+
+# Trust workdir
+RUN git config --global --add safe.directory /greptimedb
+
+# Install Rust.
+SHELL ["/bin/bash", "-c"]
+RUN curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh -s -- --no-modify-path --default-toolchain none -y
+ENV PATH /root/.cargo/bin/:$PATH
+
+# Add android toolchains
+ARG RUST_TOOLCHAIN
+RUN rustup toolchain install ${RUST_TOOLCHAIN}
+RUN rustup target add aarch64-linux-android
+
+# Install cargo-ndk
+RUN cargo install cargo-ndk
+ENV ANDROID_NDK_HOME $NDK_ROOT
+
+# Builder entrypoint.
+CMD ["cargo", "ndk", "--platform", "23", "-t", "aarch64-linux-android", "build", "--bin", "greptime", "--profile", "release", "--no-default-features"]

docker/dev-builder/ubuntu/Dockerfile

@@ -1,8 +1,13 @@
-FROM ubuntu:22.04
+FROM ubuntu:20.04
 
 ENV LANG en_US.utf8
 WORKDIR /greptimedb
 
+# Add PPA for Python 3.10.
+RUN apt-get update && \
+    DEBIAN_FRONTEND=noninteractive apt-get install -y software-properties-common && \
+    add-apt-repository ppa:deadsnakes/ppa -y
+
 # Install dependencies.
 RUN apt-get update && DEBIAN_FRONTEND=noninteractive apt-get install -y \
     libssl-dev \
@@ -14,8 +19,13 @@ RUN apt-get update && DEBIAN_FRONTEND=noninteractive apt-get install -y \
     build-essential \
     pkg-config \
     python3.10 \
-    python3.10-dev \
-    python3-pip
+    python3.10-dev
+
+# Remove Python 3.8 and install pip.
+RUN apt-get -y purge python3.8 && \
+    apt-get -y autoremove && \
+    ln -s /usr/bin/python3.10 /usr/bin/python3 && \
+    curl -sS https://bootstrap.pypa.io/get-pip.py | python3.10
 
 RUN git config --global --add safe.directory /greptimedb
 

docs/benchmarks/tsbs/v0.4.0.md

@@ -0,0 +1,61 @@
+# TSBS benchmark - v0.4.0
+
+## Environment
+
+### Local
+|        |                                    |
+| ------ | ---------------------------------- |
+| CPU    | AMD Ryzen 7 7735HS (8 core 3.2GHz) |
+| Memory | 32GB                               |
+| Disk   | SOLIDIGM SSDPFKNU010TZ             |
+| OS     | Ubuntu 22.04.2 LTS                 |
+
+### Aliyun amd64
+
+|         |                |
+| ------- | -------------- |
+| Machine | ecs.g7.4xlarge |
+| CPU     | 16 core        |
+| Memory  | 64GB           |
+| Disk    | 100G           |
+| OS      | Ubuntu  22.04  |
+
+### Aliyun arm64
+
+|         |                   |
+| ------- | ----------------- |
+| Machine | ecs.g8y.4xlarge   |
+| CPU     | 16 core           |
+| Memory  | 64GB              |
+| Disk    | 100G              |
+| OS      | Ubuntu  22.04 ARM |
+
+
+## Write performance
+
+| Environment        | Ingest rate（rows/s） |
+| ------------------ | --------------------- |
+| Local              | 365280.60             |
+| Aliyun g7.4xlarge  | 341368.72             |
+| Aliyun g8y.4xlarge | 320907.29             |
+
+
+## Query performance
+
+| Query type            | Local (ms) | Aliyun g7.4xlarge (ms) | Aliyun g8y.4xlarge (ms) |
+| --------------------- | ---------- | ---------------------- | ----------------------- |
+| cpu-max-all-1         | 50.70      | 31.46                  | 47.61                   |
+| cpu-max-all-8         | 262.16     | 129.26                 | 152.43                  |
+| double-groupby-1      | 2512.71    | 1408.19                | 1586.10                 |
+| double-groupby-5      | 3896.15    | 2304.29                | 2585.29                 |
+| double-groupby-all    | 5404.67    | 3337.61                | 3773.91                 |
+| groupby-orderby-limit | 3786.98    | 2065.72                | 2312.57                 |
+| high-cpu-1            | 71.96      | 37.29                  | 54.01                   |
+| high-cpu-all          | 9468.75    | 7595.69                | 8467.46                 |
+| lastpoint             | 13379.43   | 11253.76               | 12949.40                |
+| single-groupby-1-1-1  | 20.72      | 12.16                  | 13.35                   |
+| single-groupby-1-1-12 | 28.53      | 15.67                  | 21.62                   |
+| single-groupby-1-8-1  | 72.23      | 37.90                  | 43.52                   |
+| single-groupby-5-1-1  | 26.75      | 15.59                  | 17.48                   |
+| single-groupby-5-1-12 | 45.41      | 22.90                  | 31.96                   |
+| single-groupby-5-8-1  | 107.96     | 59.76                  | 69.58                   |

src/api/Cargo.toml

@@ -7,6 +7,7 @@ license.workspace = true
 [dependencies]
 common-base = { workspace = true }
 common-error = { workspace = true }
+common-macro = { workspace = true }
 common-time = { workspace = true }
 datatypes = { workspace = true }
 greptime-proto.workspace = true

src/api/src/error.rs

@@ -16,14 +16,16 @@ use std::any::Any;
 
 use common_error::ext::ErrorExt;
 use common_error::status_code::StatusCode;
+use common_macro::stack_trace_debug;
 use datatypes::prelude::ConcreteDataType;
 use snafu::prelude::*;
 use snafu::Location;
 
 pub type Result<T> = std::result::Result<T, Error>;
 
-#[derive(Debug, Snafu)]
+#[derive(Snafu)]
 #[snafu(visibility(pub))]
+#[stack_trace_debug]
 pub enum Error {
     #[snafu(display("Unknown proto column datatype: {}", datatype))]
     UnknownColumnDataType { datatype: i32, location: Location },

src/auth/Cargo.toml

@@ -4,8 +4,6 @@ version.workspace = true
 edition.workspace = true
 license.workspace = true
 
-# See more keys and their definitions at https://doc.rust-lang.org/cargo/reference/manifest.html
-
 [features]
 default = []
 testing = []
@@ -14,6 +12,7 @@ testing = []
 api.workspace = true
 async-trait.workspace = true
 common-error.workspace = true
+common-macro.workspace = true
 digest = "0.10"
 hex = { version = "0.4" }
 secrecy = { version = "0.8", features = ["serde", "alloc"] }

src/auth/src/error.rs

@@ -14,10 +14,12 @@
 
 use common_error::ext::{BoxedError, ErrorExt};
 use common_error::status_code::StatusCode;
+use common_macro::stack_trace_debug;
 use snafu::{Location, Snafu};
 
-#[derive(Debug, Snafu)]
+#[derive(Snafu)]
 #[snafu(visibility(pub))]
+#[stack_trace_debug]
 pub enum Error {
     #[snafu(display("Invalid config value: {}, {}", value, msg))]
     InvalidConfig { value: String, msg: String },
@@ -30,7 +32,8 @@ pub enum Error {
 
     #[snafu(display("IO error"))]
     Io {
-        source: std::io::Error,
+        #[snafu(source)]
+        error: std::io::Error,
         location: Location,
     },
 

src/catalog/Cargo.toml

@@ -16,6 +16,7 @@ async-trait = "0.1"
 common-catalog = { workspace = true }
 common-error = { workspace = true }
 common-grpc = { workspace = true }
+common-macro = { workspace = true }
 common-meta = { workspace = true }
 common-query = { workspace = true }
 common-recordbatch = { workspace = true }
@@ -30,7 +31,7 @@ futures-util.workspace = true
 lazy_static.workspace = true
 meta-client = { workspace = true }
 metrics.workspace = true
-moka = { version = "0.11", features = ["future"] }
+moka = { workspace = true, features = ["future"] }
 parking_lot = "0.12"
 partition.workspace = true
 regex.workspace = true

src/catalog/src/error.rs

@@ -17,14 +17,16 @@ use std::fmt::Debug;
 
 use common_error::ext::{BoxedError, ErrorExt};
 use common_error::status_code::StatusCode;
+use common_macro::stack_trace_debug;
 use datafusion::error::DataFusionError;
 use datatypes::prelude::ConcreteDataType;
 use snafu::{Location, Snafu};
 use table::metadata::TableId;
 use tokio::task::JoinError;
 
-#[derive(Debug, Snafu)]
+#[derive(Snafu)]
 #[snafu(visibility(pub))]
+#[stack_trace_debug]
 pub enum Error {
     #[snafu(display("Failed to list catalogs"))]
     ListCatalogs {
@@ -92,7 +94,8 @@ pub enum Error {
 
     #[snafu(display("Failed to deserialize value"))]
     ValueDeserialize {
-        source: serde_json::error::Error,
+        #[snafu(source)]
+        error: serde_json::error::Error,
         location: Location,
     },
 
@@ -142,7 +145,10 @@ pub enum Error {
     },
 
     #[snafu(display("Failed to open table in parallel"))]
-    ParallelOpenTable { source: JoinError },
+    ParallelOpenTable {
+        #[snafu(source)]
+        error: JoinError,
+    },
 
     #[snafu(display("Table not found while opening table, table info: {}", table_info))]
     TableNotFound {
@@ -210,10 +216,10 @@ pub enum Error {
     #[snafu(display("Illegal access to catalog: {} and schema: {}", catalog, schema))]
     QueryAccessDenied { catalog: String, schema: String },
 
-    #[snafu(display("msg: {}", msg))]
+    #[snafu(display(""))]
     Datafusion {
-        msg: String,
-        source: DataFusionError,
+        #[snafu(source)]
+        error: DataFusionError,
         location: Location,
     },
 

src/catalog/src/kvbackend/manager.rs

@@ -18,9 +18,7 @@ use std::sync::{Arc, Weak};
 
 use common_catalog::consts::{DEFAULT_SCHEMA_NAME, INFORMATION_SCHEMA_NAME, NUMBERS_TABLE_ID};
 use common_error::ext::BoxedError;
-use common_meta::cache_invalidator::{
-    CacheInvalidator, Context, KvCacheInvalidatorRef, TableMetadataCacheInvalidator,
-};
+use common_meta::cache_invalidator::{CacheInvalidator, CacheInvalidatorRef, Context};
 use common_meta::datanode_manager::DatanodeManagerRef;
 use common_meta::error::Result as MetaResult;
 use common_meta::key::catalog_name::CatalogNameKey;
@@ -54,7 +52,7 @@ pub struct KvBackendCatalogManager {
     // TODO(LFC): Maybe use a real implementation for Standalone mode.
     // Now we use `NoopKvCacheInvalidator` for Standalone mode. In Standalone mode, the KV backend
     // is implemented by RaftEngine. Maybe we need a cache for it?
-    table_metadata_cache_invalidator: TableMetadataCacheInvalidator,
+    cache_invalidator: CacheInvalidatorRef,
     partition_manager: PartitionRuleManagerRef,
     table_metadata_manager: TableMetadataManagerRef,
     datanode_manager: DatanodeManagerRef,
@@ -65,13 +63,13 @@ pub struct KvBackendCatalogManager {
 #[async_trait::async_trait]
 impl CacheInvalidator for KvBackendCatalogManager {
     async fn invalidate_table_name(&self, ctx: &Context, table_name: TableName) -> MetaResult<()> {
-        self.table_metadata_cache_invalidator
+        self.cache_invalidator
             .invalidate_table_name(ctx, table_name)
             .await
     }
 
     async fn invalidate_table_id(&self, ctx: &Context, table_id: TableId) -> MetaResult<()> {
-        self.table_metadata_cache_invalidator
+        self.cache_invalidator
             .invalidate_table_id(ctx, table_id)
             .await
     }
@@ -80,15 +78,13 @@ impl CacheInvalidator for KvBackendCatalogManager {
 impl KvBackendCatalogManager {
     pub fn new(
         backend: KvBackendRef,
-        backend_cache_invalidator: KvCacheInvalidatorRef,
+        cache_invalidator: CacheInvalidatorRef,
         datanode_manager: DatanodeManagerRef,
     ) -> Arc<Self> {
         Arc::new_cyclic(|me| Self {
             partition_manager: Arc::new(PartitionRuleManager::new(backend.clone())),
             table_metadata_manager: Arc::new(TableMetadataManager::new(backend)),
-            table_metadata_cache_invalidator: TableMetadataCacheInvalidator::new(
-                backend_cache_invalidator.clone(),
-            ),
+            cache_invalidator,
             datanode_manager,
             system_catalog: SystemCatalog {
                 catalog_manager: me.clone(),
@@ -107,12 +103,6 @@ impl KvBackendCatalogManager {
     pub fn datanode_manager(&self) -> DatanodeManagerRef {
         self.datanode_manager.clone()
     }
-
-    pub async fn invalidate_schema(&self, catalog: &str, schema: &str) {
-        self.table_metadata_cache_invalidator
-            .invalidate_schema(catalog, schema)
-            .await
-    }
 }
 
 #[async_trait::async_trait]
@@ -229,6 +219,7 @@ impl CatalogManager for KvBackendCatalogManager {
             .get(table_id)
             .await
             .context(TableMetadataManagerSnafu)?
+            .map(|v| v.into_inner())
         else {
             return Ok(None);
         };

src/client/Cargo.toml

@@ -16,6 +16,7 @@ common-base = { workspace = true }
 common-catalog = { workspace = true }
 common-error = { workspace = true }
 common-grpc = { workspace = true }
+common-macro = { workspace = true }
 common-meta = { workspace = true }
 common-query = { workspace = true }
 common-recordbatch = { workspace = true }
@@ -26,7 +27,7 @@ datatypes = { workspace = true }
 derive_builder.workspace = true
 enum_dispatch = "0.3"
 futures-util.workspace = true
-moka = { version = "0.9", features = ["future"] }
+moka = { workspace = true, features = ["future"] }
 parking_lot = "0.12"
 prost.workspace = true
 rand.workspace = true

src/client/examples/stream_ingest.rs

@@ -42,14 +42,14 @@ async fn run() {
         .insert(vec![to_insert_request(weather_records_1())])
         .await
     {
-        error!("Error: {e}");
+        error!("Error: {e:?}");
     }
 
     if let Err(e) = stream_inserter
         .insert(vec![to_insert_request(weather_records_2())])
         .await
     {
-        error!("Error: {e}");
+        error!("Error: {e:?}");
     }
 
     let result = stream_inserter.finish().await;
@@ -59,7 +59,7 @@ async fn run() {
             info!("Rows written: {rows}");
         }
         Err(e) => {
-            error!("Error: {e}");
+            error!("Error: {e:?}");
         }
     };
 }

src/client/src/client.rs

@@ -139,11 +139,19 @@ impl Client {
     }
 
     fn max_grpc_recv_message_size(&self) -> usize {
-        self.inner.channel_manager.config().max_recv_message_size
+        self.inner
+            .channel_manager
+            .config()
+            .max_recv_message_size
+            .as_bytes() as usize
     }
 
     fn max_grpc_send_message_size(&self) -> usize {
-        self.inner.channel_manager.config().max_send_message_size
+        self.inner
+            .channel_manager
+            .config()
+            .max_send_message_size
+            .as_bytes() as usize
     }
 
     pub(crate) fn make_flight_client(&self) -> Result<FlightClient> {

src/client/src/database.rs

@@ -167,11 +167,14 @@ impl Database {
         }
     }
 
-    pub async fn sql(&self, sql: &str) -> Result<Output> {
+    pub async fn sql<S>(&self, sql: S) -> Result<Output>
+    where
+        S: AsRef<str>,
+    {
         let _timer = timer!(metrics::METRIC_GRPC_SQL);
         self.do_get(
             Request::Query(QueryRequest {
-                query: Some(Query::Sql(sql.to_string())),
+                query: Some(Query::Sql(sql.as_ref().to_string())),
             }),
             0,
         )
@@ -276,7 +279,7 @@ impl Database {
                 source: BoxedError::new(ServerSnafu { code, msg }.build()),
             };
             logging::error!(
-                "Failed to do Flight get, addr: {}, code: {}, source: {}",
+                "Failed to do Flight get, addr: {}, code: {}, source: {:?}",
                 client.addr(),
                 tonic_code,
                 error

src/client/src/error.rs

@@ -17,11 +17,13 @@ use std::any::Any;
 use common_error::ext::{BoxedError, ErrorExt};
 use common_error::status_code::StatusCode;
 use common_error::{GREPTIME_ERROR_CODE, GREPTIME_ERROR_MSG};
+use common_macro::stack_trace_debug;
 use snafu::{Location, Snafu};
 use tonic::{Code, Status};
 
-#[derive(Debug, Snafu)]
+#[derive(Snafu)]
 #[snafu(visibility(pub))]
+#[stack_trace_debug]
 pub enum Error {
     #[snafu(display("Illegal Flight messages, reason: {}", reason))]
     IllegalFlightMessages { reason: String, location: Location },

src/client/src/lib.rs

@@ -26,6 +26,8 @@ use api::v1::greptime_response::Response;
 use api::v1::{AffectedRows, GreptimeResponse};
 pub use common_catalog::consts::{DEFAULT_CATALOG_NAME, DEFAULT_SCHEMA_NAME};
 use common_error::status_code::StatusCode;
+pub use common_query::Output;
+pub use common_recordbatch::{RecordBatches, SendableRecordBatchStream};
 use snafu::OptionExt;
 
 pub use self::client::Client;

src/cmd/Cargo.toml

@@ -26,6 +26,7 @@ common-base = { 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 }
@@ -48,6 +49,7 @@ metrics.workspace = true
 mito2 = { workspace = true }
 nu-ansi-term = "0.46"
 partition = { workspace = true }
+plugins.workspace = true
 prost.workspace = true
 query = { workspace = true }
 rand.workspace = true

src/cmd/src/cli.rs

@@ -14,6 +14,7 @@
 
 mod bench;
 mod cmd;
+mod export;
 mod helper;
 mod repl;
 // TODO(weny): Removes it
@@ -27,6 +28,7 @@ use common_telemetry::logging::LoggingOptions;
 pub use repl::Repl;
 use upgrade::UpgradeCommand;
 
+use self::export::ExportCommand;
 use crate::error::Result;
 use crate::options::{Options, TopLevelOptions};
 
@@ -78,17 +80,19 @@ impl Command {
 
 #[derive(Parser)]
 enum SubCommand {
-    Attach(AttachCommand),
+    // Attach(AttachCommand),
     Upgrade(UpgradeCommand),
     Bench(BenchTableMetadataCommand),
+    Export(ExportCommand),
 }
 
 impl SubCommand {
     async fn build(self) -> Result<Instance> {
         match self {
-            SubCommand::Attach(cmd) => cmd.build().await,
+            // SubCommand::Attach(cmd) => cmd.build().await,
             SubCommand::Upgrade(cmd) => cmd.build().await,
             SubCommand::Bench(cmd) => cmd.build().await,
+            SubCommand::Export(cmd) => cmd.build().await,
         }
     }
 }
@@ -104,51 +108,9 @@ pub(crate) struct AttachCommand {
 }
 
 impl AttachCommand {
+    #[allow(dead_code)]
     async fn build(self) -> Result<Instance> {
         let repl = Repl::try_new(&self).await?;
         Ok(Instance::Repl(repl))
     }
 }
-
-#[cfg(test)]
-mod tests {
-    use super::*;
-
-    #[test]
-    fn test_load_options() {
-        let cmd = Command {
-            cmd: SubCommand::Attach(AttachCommand {
-                grpc_addr: String::from(""),
-                meta_addr: None,
-                disable_helper: false,
-            }),
-        };
-
-        let opts = cmd.load_options(TopLevelOptions::default()).unwrap();
-        let logging_opts = opts.logging_options();
-        assert_eq!("/tmp/greptimedb/logs", logging_opts.dir);
-        assert!(logging_opts.level.is_none());
-        assert!(!logging_opts.enable_jaeger_tracing);
-    }
-
-    #[test]
-    fn test_top_level_options() {
-        let cmd = Command {
-            cmd: SubCommand::Attach(AttachCommand {
-                grpc_addr: String::from(""),
-                meta_addr: None,
-                disable_helper: false,
-            }),
-        };
-
-        let opts = cmd
-            .load_options(TopLevelOptions {
-                log_dir: Some("/tmp/greptimedb/test/logs".to_string()),
-                log_level: Some("debug".to_string()),
-            })
-            .unwrap();
-        let logging_opts = opts.logging_options();
-        assert_eq!("/tmp/greptimedb/test/logs", logging_opts.dir);
-        assert_eq!("debug", logging_opts.level.as_ref().unwrap());
-    }
-}

src/cmd/src/cli/export.rs

@@ -0,0 +1,395 @@
+// 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::path::Path;
+use std::sync::Arc;
+
+use async_trait::async_trait;
+use clap::{Parser, ValueEnum};
+use client::{Client, Database, DEFAULT_SCHEMA_NAME};
+use common_query::Output;
+use common_recordbatch::util::collect;
+use common_telemetry::{debug, error, info, warn};
+use datatypes::scalars::ScalarVector;
+use datatypes::vectors::{StringVector, Vector};
+use snafu::{OptionExt, ResultExt};
+use tokio::fs::File;
+use tokio::io::AsyncWriteExt;
+use tokio::sync::Semaphore;
+
+use crate::cli::{Instance, Tool};
+use crate::error::{
+    CollectRecordBatchesSnafu, ConnectServerSnafu, EmptyResultSnafu, Error, FileIoSnafu,
+    InvalidDatabaseNameSnafu, NotDataFromOutputSnafu, RequestDatabaseSnafu, Result,
+};
+
+type TableReference = (String, String, String);
+
+#[derive(Debug, Default, Clone, ValueEnum)]
+enum ExportTarget {
+    /// Corresponding to `SHOW CREATE TABLE`
+    #[default]
+    CreateTable,
+    /// Corresponding to `EXPORT TABLE`
+    TableData,
+}
+
+#[derive(Debug, Default, Parser)]
+pub struct ExportCommand {
+    /// Server address to connect
+    #[clap(long)]
+    addr: String,
+
+    /// Directory to put the exported data. E.g.: /tmp/greptimedb-export
+    #[clap(long)]
+    output_dir: String,
+
+    /// The name of the catalog to export. Default to "greptime-*"".
+    #[clap(long, default_value = "")]
+    database: String,
+
+    /// Parallelism of the export.
+    #[clap(long, short = 'j', default_value = "1")]
+    export_jobs: usize,
+
+    /// Max retry times for each job.
+    #[clap(long, default_value = "3")]
+    max_retry: usize,
+
+    /// Things to export
+    #[clap(long, short = 't', value_enum)]
+    target: ExportTarget,
+}
+
+impl ExportCommand {
+    pub async fn build(&self) -> Result<Instance> {
+        let client = Client::with_urls([self.addr.clone()]);
+        client
+            .health_check()
+            .await
+            .with_context(|_| ConnectServerSnafu {
+                addr: self.addr.clone(),
+            })?;
+        let (catalog, schema) = split_database(&self.database)?;
+        let database_client = Database::new(
+            catalog.clone(),
+            schema.clone().unwrap_or(DEFAULT_SCHEMA_NAME.to_string()),
+            client,
+        );
+
+        Ok(Instance::Tool(Box::new(Export {
+            client: database_client,
+            catalog,
+            schema,
+            output_dir: self.output_dir.clone(),
+            parallelism: self.export_jobs,
+            target: self.target.clone(),
+        })))
+    }
+}
+
+pub struct Export {
+    client: Database,
+    catalog: String,
+    schema: Option<String>,
+    output_dir: String,
+    parallelism: usize,
+    target: ExportTarget,
+}
+
+impl Export {
+    /// Iterate over all db names.
+    ///
+    /// Newbie: `db_name` is catalog + schema.
+    async fn iter_db_names(&self) -> Result<Vec<(String, String)>> {
+        if let Some(schema) = &self.schema {
+            Ok(vec![(self.catalog.clone(), schema.clone())])
+        } else {
+            let mut client = self.client.clone();
+            client.set_catalog(self.catalog.clone());
+            let result =
+                client
+                    .sql("show databases")
+                    .await
+                    .with_context(|_| RequestDatabaseSnafu {
+                        sql: "show databases".to_string(),
+                    })?;
+            let Output::Stream(stream) = result else {
+                NotDataFromOutputSnafu.fail()?
+            };
+            let record_batch = collect(stream)
+                .await
+                .context(CollectRecordBatchesSnafu)?
+                .pop()
+                .context(EmptyResultSnafu)?;
+            let schemas = record_batch
+                .column(0)
+                .as_any()
+                .downcast_ref::<StringVector>()
+                .unwrap();
+            let mut result = Vec::with_capacity(schemas.len());
+            for i in 0..schemas.len() {
+                let schema = schemas.get_data(i).unwrap().to_owned();
+                result.push((self.catalog.clone(), schema));
+            }
+            Ok(result)
+        }
+    }
+
+    /// Return a list of [`TableReference`] to be exported.
+    /// Includes all tables under the given `catalog` and `schema`
+    async fn get_table_list(&self, catalog: &str, schema: &str) -> Result<Vec<TableReference>> {
+        // TODO: SQL injection hurts
+        let sql = format!(
+            "select table_catalog, table_schema, table_name from \
+            information_schema.tables where table_type = \'BASE TABLE\'\
+            and table_catalog = \'{catalog}\' and table_schema = \'{schema}\'",
+        );
+        let mut client = self.client.clone();
+        client.set_catalog(catalog);
+        client.set_schema(schema);
+        let result = client
+            .sql(&sql)
+            .await
+            .with_context(|_| RequestDatabaseSnafu { sql })?;
+        let Output::Stream(stream) = result else {
+            NotDataFromOutputSnafu.fail()?
+        };
+        let Some(record_batch) = collect(stream)
+            .await
+            .context(CollectRecordBatchesSnafu)?
+            .pop()
+        else {
+            return Ok(vec![]);
+        };
+
+        debug!("Fetched table list: {}", record_batch.pretty_print());
+
+        if record_batch.num_rows() == 0 {
+            return Ok(vec![]);
+        }
+
+        let mut result = Vec::with_capacity(record_batch.num_rows());
+        let catalog_column = record_batch
+            .column(0)
+            .as_any()
+            .downcast_ref::<StringVector>()
+            .unwrap();
+        let schema_column = record_batch
+            .column(1)
+            .as_any()
+            .downcast_ref::<StringVector>()
+            .unwrap();
+        let table_column = record_batch
+            .column(2)
+            .as_any()
+            .downcast_ref::<StringVector>()
+            .unwrap();
+        for i in 0..record_batch.num_rows() {
+            let catalog = catalog_column.get_data(i).unwrap().to_owned();
+            let schema = schema_column.get_data(i).unwrap().to_owned();
+            let table = table_column.get_data(i).unwrap().to_owned();
+            result.push((catalog, schema, table));
+        }
+
+        Ok(result)
+    }
+
+    async fn show_create_table(&self, catalog: &str, schema: &str, table: &str) -> Result<String> {
+        let sql = format!("show create table {}.{}.{}", catalog, schema, table);
+        let mut client = self.client.clone();
+        client.set_catalog(catalog);
+        client.set_schema(schema);
+        let result = client
+            .sql(&sql)
+            .await
+            .with_context(|_| RequestDatabaseSnafu { sql })?;
+        let Output::Stream(stream) = result else {
+            NotDataFromOutputSnafu.fail()?
+        };
+        let record_batch = collect(stream)
+            .await
+            .context(CollectRecordBatchesSnafu)?
+            .pop()
+            .context(EmptyResultSnafu)?;
+        let create_table = record_batch
+            .column(1)
+            .as_any()
+            .downcast_ref::<StringVector>()
+            .unwrap()
+            .get_data(0)
+            .unwrap();
+
+        Ok(format!("{create_table};\n"))
+    }
+
+    async fn export_create_table(&self) -> Result<()> {
+        let semaphore = Arc::new(Semaphore::new(self.parallelism));
+        let db_names = self.iter_db_names().await?;
+        let db_count = db_names.len();
+        let mut tasks = Vec::with_capacity(db_names.len());
+        for (catalog, schema) in db_names {
+            let semaphore_moved = semaphore.clone();
+            tasks.push(async move {
+                let _permit = semaphore_moved.acquire().await.unwrap();
+                let table_list = self.get_table_list(&catalog, &schema).await?;
+                let table_count = table_list.len();
+                tokio::fs::create_dir_all(&self.output_dir)
+                    .await
+                    .context(FileIoSnafu)?;
+                let output_file =
+                    Path::new(&self.output_dir).join(format!("{catalog}-{schema}.sql"));
+                let mut file = File::create(output_file).await.context(FileIoSnafu)?;
+                for (c, s, t) in table_list {
+                    match self.show_create_table(&c, &s, &t).await {
+                        Err(e) => {
+                            error!(e; "Failed to export table {}.{}.{}", c, s, t)
+                        }
+                        Ok(create_table) => {
+                            file.write_all(create_table.as_bytes())
+                                .await
+                                .context(FileIoSnafu)?;
+                        }
+                    }
+                }
+                info!("finished exporting {catalog}.{schema} with {table_count} tables",);
+                Ok::<(), Error>(())
+            });
+        }
+
+        let success = futures::future::join_all(tasks)
+            .await
+            .into_iter()
+            .filter(|r| match r {
+                Ok(_) => true,
+                Err(e) => {
+                    error!(e; "export job failed");
+                    false
+                }
+            })
+            .count();
+
+        info!("success {success}/{db_count} jobs");
+
+        Ok(())
+    }
+
+    async fn export_table_data(&self) -> Result<()> {
+        let semaphore = Arc::new(Semaphore::new(self.parallelism));
+        let db_names = self.iter_db_names().await?;
+        let db_count = db_names.len();
+        let mut tasks = Vec::with_capacity(db_names.len());
+        for (catalog, schema) in db_names {
+            let semaphore_moved = semaphore.clone();
+            tasks.push(async move {
+                let _permit = semaphore_moved.acquire().await.unwrap();
+                tokio::fs::create_dir_all(&self.output_dir)
+                    .await
+                    .context(FileIoSnafu)?;
+                let output_dir = Path::new(&self.output_dir).join(format!("{catalog}-{schema}/"));
+
+                let mut client = self.client.clone();
+                client.set_catalog(catalog.clone());
+                client.set_schema(schema.clone());
+
+                // copy database to
+                let sql = format!(
+                    "copy database {} to '{}' with (format='parquet');",
+                    schema,
+                    output_dir.to_str().unwrap()
+                );
+                client
+                    .sql(sql.clone())
+                    .await
+                    .context(RequestDatabaseSnafu { sql })?;
+                info!("finished exporting {catalog}.{schema} data");
+
+                // export copy from sql
+                let dir_filenames = match output_dir.read_dir() {
+                    Ok(dir) => dir,
+                    Err(_) => {
+                        warn!("empty database {catalog}.{schema}");
+                        return Ok(());
+                    }
+                };
+
+                let copy_from_file =
+                    Path::new(&self.output_dir).join(format!("{catalog}-{schema}_copy_from.sql"));
+                let mut file = File::create(copy_from_file).await.context(FileIoSnafu)?;
+
+                let copy_from_sql = dir_filenames
+                    .into_iter()
+                    .map(|file| {
+                        let file = file.unwrap();
+                        let filename = file.file_name().into_string().unwrap();
+
+                        format!(
+                            "copy {} from '{}' with (format='parquet');\n",
+                            filename.replace(".parquet", ""),
+                            file.path().to_str().unwrap()
+                        )
+                    })
+                    .collect::<Vec<_>>()
+                    .join("");
+                file.write_all(copy_from_sql.as_bytes())
+                    .await
+                    .context(FileIoSnafu)?;
+
+                info!("finished exporting {catalog}.{schema} copy_from.sql");
+
+                Ok::<(), Error>(())
+            });
+        }
+
+        let success = futures::future::join_all(tasks)
+            .await
+            .into_iter()
+            .filter(|r| match r {
+                Ok(_) => true,
+                Err(e) => {
+                    error!(e; "export job failed");
+                    false
+                }
+            })
+            .count();
+
+        info!("success {success}/{db_count} jobs");
+
+        Ok(())
+    }
+}
+
+#[async_trait]
+impl Tool for Export {
+    async fn do_work(&self) -> Result<()> {
+        match self.target {
+            ExportTarget::CreateTable => self.export_create_table().await,
+            ExportTarget::TableData => self.export_table_data().await,
+        }
+    }
+}
+
+/// Split at `-`.
+fn split_database(database: &str) -> Result<(String, Option<String>)> {
+    let (catalog, schema) = database
+        .split_once('-')
+        .with_context(|| InvalidDatabaseNameSnafu {
+            database: database.to_string(),
+        })?;
+    if schema == "*" {
+        Ok((catalog.to_string(), None))
+    } else {
+        Ok((catalog.to_string(), Some(schema.to_string())))
+    }
+}

src/cmd/src/cli/repl.rs

@@ -35,7 +35,7 @@ use query::QueryEngine;
 use rustyline::error::ReadlineError;
 use rustyline::Editor;
 use session::context::QueryContext;
-use snafu::{ErrorCompat, ResultExt};
+use snafu::ResultExt;
 use substrait::{DFLogicalSubstraitConvertor, SubstraitPlan};
 
 use crate::cli::cmd::ReplCommand;
@@ -148,7 +148,7 @@ impl Repl {
             .await
             .map_err(|e| {
                 let status_code = e.status_code();
-                let root_cause = e.iter_chain().last().unwrap();
+                let root_cause = e.output_msg();
                 println!("Error: {}({status_code}), {root_cause}", status_code as u32)
             })
             .is_ok()
@@ -257,10 +257,11 @@ async fn create_query_engine(meta_addr: &str) -> Result<DatafusionQueryEngine> {
         cached_meta_backend.clone(),
         datanode_clients,
     );
-    let plugins: Arc<Plugins> = Default::default();
+    let plugins: Plugins = Default::default();
     let state = Arc::new(QueryEngineState::new(
         catalog_list,
         None,
+        None,
         false,
         plugins.clone(),
     ));

src/cmd/src/cli/upgrade.rs

@@ -20,7 +20,7 @@ use client::api::v1::meta::TableRouteValue;
 use common_meta::ddl::utils::region_storage_path;
 use common_meta::error as MetaError;
 use common_meta::key::catalog_name::{CatalogNameKey, CatalogNameValue};
-use common_meta::key::datanode_table::{DatanodeTableKey, DatanodeTableValue};
+use common_meta::key::datanode_table::{DatanodeTableKey, DatanodeTableValue, RegionInfo};
 use common_meta::key::schema_name::{SchemaNameKey, SchemaNameValue};
 use common_meta::key::table_info::{TableInfoKey, TableInfoValue};
 use common_meta::key::table_name::{TableNameKey, TableNameValue};
@@ -405,8 +405,11 @@ impl MigrateTableMetadata {
                     DatanodeTableValue::new(
                         table_id,
                         regions,
-                        engine.to_string(),
-                        region_storage_path.clone(),
+                        RegionInfo {
+                            engine: engine.to_string(),
+                            region_storage_path: region_storage_path.clone(),
+                            region_options: (&value.table_info.meta.options).into(),
+                        },
                     ),
                 )
             })

src/cmd/src/datanode.rs

@@ -31,6 +31,10 @@ pub struct Instance {
 
 impl Instance {
     pub async fn start(&mut self) -> Result<()> {
+        plugins::start_datanode_plugins(self.datanode.plugins())
+            .await
+            .context(StartDatanodeSnafu)?;
+
         self.datanode.start().await.context(StartDatanodeSnafu)
     }
 
@@ -92,6 +96,8 @@ struct StartCommand {
     #[clap(long)]
     data_home: Option<String>,
     #[clap(long)]
+    wal_dir: Option<String>,
+    #[clap(long)]
     http_addr: Option<String>,
     #[clap(long)]
     http_timeout: Option<u64>,
@@ -145,6 +151,10 @@ impl StartCommand {
             opts.storage.data_home = data_home.clone();
         }
 
+        if let Some(wal_dir) = &self.wal_dir {
+            opts.wal.dir = Some(wal_dir.clone());
+        }
+
         if let Some(http_addr) = &self.http_addr {
             opts.http.addr = http_addr.clone();
         }
@@ -159,11 +169,15 @@ impl StartCommand {
         Ok(Options::Datanode(Box::new(opts)))
     }
 
-    async fn build(self, opts: DatanodeOptions) -> Result<Instance> {
+    async fn build(self, mut opts: DatanodeOptions) -> Result<Instance> {
+        let plugins = plugins::setup_datanode_plugins(&mut opts)
+            .await
+            .context(StartDatanodeSnafu)?;
+
         logging::info!("Datanode start command: {:#?}", self);
         logging::info!("Datanode options: {:#?}", opts);
 
-        let datanode = DatanodeBuilder::new(opts, None, Default::default())
+        let datanode = DatanodeBuilder::new(opts, None, plugins)
             .build()
             .await
             .context(StartDatanodeSnafu)?;
@@ -180,6 +194,7 @@ mod tests {
     use common_base::readable_size::ReadableSize;
     use common_test_util::temp_dir::create_named_temp_file;
     use datanode::config::{CompactionConfig, FileConfig, ObjectStoreConfig, RegionManifestConfig};
+    use servers::heartbeat_options::HeartbeatOptions;
     use servers::Mode;
 
     use super::*;
@@ -196,11 +211,14 @@ mod tests {
             rpc_hostname = "127.0.0.1"
             rpc_runtime_size = 8
 
+            [heartbeat]
+            interval = "300ms"
+
             [meta_client]
             metasrv_addrs = ["127.0.0.1:3002"]
-            timeout_millis = 3000
-            connect_timeout_millis = 5000
-            ddl_timeout_millis= 10000
+            timeout = "3s"
+            connect_timeout = "5s"
+            ddl_timeout = "10s"
             tcp_nodelay = true
 
             [wal]
@@ -243,24 +261,33 @@ mod tests {
 
         assert_eq!("127.0.0.1:3001".to_string(), options.rpc_addr);
         assert_eq!(Some(42), options.node_id);
+        assert_eq!("/other/wal", options.wal.dir.unwrap());
 
         assert_eq!(Duration::from_secs(600), options.wal.purge_interval);
         assert_eq!(1024 * 1024 * 1024, options.wal.file_size.0);
         assert_eq!(1024 * 1024 * 1024 * 50, options.wal.purge_threshold.0);
         assert!(!options.wal.sync_write);
 
+        let HeartbeatOptions {
+            interval: heart_beat_interval,
+            ..
+        } = options.heartbeat;
+
+        assert_eq!(300, heart_beat_interval.as_millis());
+
         let MetaClientOptions {
             metasrv_addrs: metasrv_addr,
-            timeout_millis,
-            connect_timeout_millis,
+            timeout,
+            connect_timeout,
+            ddl_timeout,
             tcp_nodelay,
-            ddl_timeout_millis,
+            ..
         } = options.meta_client.unwrap();
 
         assert_eq!(vec!["127.0.0.1:3002".to_string()], metasrv_addr);
-        assert_eq!(5000, connect_timeout_millis);
-        assert_eq!(10000, ddl_timeout_millis);
-        assert_eq!(3000, timeout_millis);
+        assert_eq!(5000, connect_timeout.as_millis());
+        assert_eq!(10000, ddl_timeout.as_millis());
+        assert_eq!(3000, timeout.as_millis());
         assert!(tcp_nodelay);
         assert_eq!("/tmp/greptimedb/", options.storage.data_home);
         assert!(matches!(
@@ -354,8 +381,8 @@ mod tests {
             rpc_runtime_size = 8
 
             [meta_client]
-            timeout_millis = 3000
-            connect_timeout_millis = 5000
+            timeout = "3s"
+            connect_timeout = "5s"
             tcp_nodelay = true
 
             [wal]
@@ -419,6 +446,7 @@ mod tests {
             || {
                 let command = StartCommand {
                     config_file: Some(file.path().to_str().unwrap().to_string()),
+                    wal_dir: Some("/other/wal/dir".to_string()),
                     env_prefix: env_prefix.to_string(),
                     ..Default::default()
                 };
@@ -446,6 +474,9 @@ mod tests {
                 // Should be read from config file, config file > env > default values.
                 assert_eq!(opts.storage.compaction.max_purge_tasks, 32);
 
+                // Should be read from cli, cli > config file > env > default values.
+                assert_eq!(opts.wal.dir.unwrap(), "/other/wal/dir");
+
                 // Should be default value.
                 assert_eq!(
                     opts.storage.manifest.checkpoint_margin,

src/cmd/src/error.rs

@@ -16,12 +16,14 @@ use std::any::Any;
 
 use common_error::ext::ErrorExt;
 use common_error::status_code::StatusCode;
+use common_macro::stack_trace_debug;
 use config::ConfigError;
 use rustyline::error::ReadlineError;
 use snafu::{Location, Snafu};
 
-#[derive(Debug, Snafu)]
+#[derive(Snafu)]
 #[snafu(visibility(pub))]
+#[stack_trace_debug]
 pub enum Error {
     #[snafu(display("Failed to create default catalog and schema"))]
     InitMetadata {
@@ -35,6 +37,18 @@ pub enum Error {
         source: common_meta::error::Error,
     },
 
+    #[snafu(display("Failed to start procedure manager"))]
+    StartProcedureManager {
+        location: Location,
+        source: common_procedure::error::Error,
+    },
+
+    #[snafu(display("Failed to stop procedure manager"))]
+    StopProcedureManager {
+        location: Location,
+        source: common_procedure::error::Error,
+    },
+
     #[snafu(display("Failed to start datanode"))]
     StartDatanode {
         location: Location,
@@ -83,12 +97,6 @@ pub enum Error {
     #[snafu(display("Illegal config: {}", msg))]
     IllegalConfig { msg: String, location: Location },
 
-    #[snafu(display("Illegal auth config"))]
-    IllegalAuthConfig {
-        location: Location,
-        source: auth::error::Error,
-    },
-
     #[snafu(display("Unsupported selector type: {}", selector_type))]
     UnsupportedSelectorType {
         selector_type: String,
@@ -101,13 +109,15 @@ pub enum Error {
 
     #[snafu(display("Cannot create REPL"))]
     ReplCreation {
-        source: ReadlineError,
+        #[snafu(source)]
+        error: ReadlineError,
         location: Location,
     },
 
     #[snafu(display("Error reading command"))]
     Readline {
-        source: ReadlineError,
+        #[snafu(source)]
+        error: ReadlineError,
         location: Location,
     },
 
@@ -157,7 +167,8 @@ pub enum Error {
 
     #[snafu(display("Failed to load layered config"))]
     LoadLayeredConfig {
-        source: ConfigError,
+        #[snafu(source)]
+        error: ConfigError,
         location: Location,
     },
 
@@ -170,9 +181,50 @@ pub enum Error {
     #[snafu(display("Failed to connect to Etcd at {etcd_addr}"))]
     ConnectEtcd {
         etcd_addr: String,
-        source: etcd_client::Error,
+        #[snafu(source)]
+        error: etcd_client::Error,
         location: Location,
     },
+
+    #[snafu(display("Failed to connect server at {addr}"))]
+    ConnectServer {
+        addr: String,
+        source: client::error::Error,
+        location: Location,
+    },
+
+    #[snafu(display("Failed to serde json"))]
+    SerdeJson {
+        #[snafu(source)]
+        error: serde_json::error::Error,
+        location: Location,
+    },
+
+    #[snafu(display("Expect data from output, but got another thing"))]
+    NotDataFromOutput { location: Location },
+
+    #[snafu(display("Empty result from output"))]
+    EmptyResult { location: Location },
+
+    #[snafu(display("Failed to manipulate file"))]
+    FileIo {
+        location: Location,
+        #[snafu(source)]
+        error: std::io::Error,
+    },
+
+    #[snafu(display("Invalid database name: {}", database))]
+    InvalidDatabaseName {
+        location: Location,
+        database: String,
+    },
+
+    #[snafu(display("Failed to create directory {}", dir))]
+    CreateDir {
+        dir: String,
+        #[snafu(source)]
+        error: std::io::Error,
+    },
 }
 
 pub type Result<T> = std::result::Result<T, Error>;
@@ -191,13 +243,18 @@ impl ErrorExt for Error {
             Error::IterStream { source, .. } | Error::InitMetadata { source, .. } => {
                 source.status_code()
             }
+            Error::ConnectServer { source, .. } => source.status_code(),
             Error::MissingConfig { .. }
             | Error::LoadLayeredConfig { .. }
             | Error::IllegalConfig { .. }
             | Error::InvalidReplCommand { .. }
-            | Error::IllegalAuthConfig { .. }
-            | Error::ConnectEtcd { .. } => StatusCode::InvalidArguments,
-
+            | Error::ConnectEtcd { .. }
+            | Error::NotDataFromOutput { .. }
+            | Error::CreateDir { .. }
+            | Error::EmptyResult { .. }
+            | Error::InvalidDatabaseName { .. } => StatusCode::InvalidArguments,
+            Error::StartProcedureManager { source, .. }
+            | Error::StopProcedureManager { source, .. } => source.status_code(),
             Error::ReplCreation { .. } | Error::Readline { .. } => StatusCode::Internal,
             Error::RequestDatabase { source, .. } => source.status_code(),
             Error::CollectRecordBatches { source, .. }
@@ -208,6 +265,8 @@ impl ErrorExt for Error {
             }
             Error::SubstraitEncodeLogicalPlan { source, .. } => source.status_code(),
             Error::StartCatalogManager { source, .. } => source.status_code(),
+
+            Error::SerdeJson { .. } | Error::FileIo { .. } => StatusCode::Unexpected,
         }
     }
 

src/cmd/src/frontend.rs

@@ -12,11 +12,9 @@
 // See the License for the specific language governing permissions and
 // limitations under the License.
 
-use std::sync::Arc;
+use std::time::Duration;
 
-use auth::UserProviderRef;
 use clap::Parser;
-use common_base::Plugins;
 use common_telemetry::logging;
 use frontend::frontend::FrontendOptions;
 use frontend::instance::{FrontendInstance, Instance as FeInstance};
@@ -25,7 +23,7 @@ use servers::tls::{TlsMode, TlsOption};
 use servers::Mode;
 use snafu::ResultExt;
 
-use crate::error::{self, IllegalAuthConfigSnafu, Result};
+use crate::error::{self, Result, StartFrontendSnafu};
 use crate::options::{Options, TopLevelOptions};
 
 pub struct Instance {
@@ -34,10 +32,11 @@ pub struct Instance {
 
 impl Instance {
     pub async fn start(&mut self) -> Result<()> {
-        self.frontend
-            .start()
+        plugins::start_frontend_plugins(self.frontend.plugins().clone())
             .await
-            .context(error::StartFrontendSnafu)
+            .context(StartFrontendSnafu)?;
+
+        self.frontend.start().await.context(StartFrontendSnafu)
     }
 
     pub async fn stop(&self) -> Result<()> {
@@ -88,7 +87,9 @@ pub struct StartCommand {
     #[clap(long)]
     http_addr: Option<String>,
     #[clap(long)]
-    grpc_addr: Option<String>,
+    http_timeout: Option<u64>,
+    #[clap(long)]
+    rpc_addr: Option<String>,
     #[clap(long)]
     mysql_addr: Option<String>,
     #[clap(long)]
@@ -141,11 +142,15 @@ impl StartCommand {
             opts.http.addr = addr.clone()
         }
 
+        if let Some(http_timeout) = self.http_timeout {
+            opts.http.timeout = Duration::from_secs(http_timeout)
+        }
+
         if let Some(disable_dashboard) = self.disable_dashboard {
             opts.http.disable_dashboard = disable_dashboard;
         }
 
-        if let Some(addr) = &self.grpc_addr {
+        if let Some(addr) = &self.rpc_addr {
             opts.grpc.addr = addr.clone()
         }
 
@@ -177,38 +182,32 @@ impl StartCommand {
             opts.mode = Mode::Distributed;
         }
 
+        opts.user_provider = self.user_provider.clone();
+
         Ok(Options::Frontend(Box::new(opts)))
     }
 
-    async fn build(self, opts: FrontendOptions) -> Result<Instance> {
+    async fn build(self, mut opts: FrontendOptions) -> Result<Instance> {
+        let plugins = plugins::setup_frontend_plugins(&mut opts)
+            .await
+            .context(StartFrontendSnafu)?;
+
         logging::info!("Frontend start command: {:#?}", self);
         logging::info!("Frontend options: {:#?}", opts);
 
-        let plugins = Arc::new(load_frontend_plugins(&self.user_provider)?);
-
         let mut instance = FeInstance::try_new_distributed(&opts, plugins.clone())
             .await
-            .context(error::StartFrontendSnafu)?;
+            .context(StartFrontendSnafu)?;
 
         instance
             .build_servers(&opts)
             .await
-            .context(error::StartFrontendSnafu)?;
+            .context(StartFrontendSnafu)?;
 
         Ok(Instance { frontend: instance })
     }
 }
 
-pub fn load_frontend_plugins(user_provider: &Option<String>) -> Result<Plugins> {
-    let plugins = Plugins::new();
-
-    if let Some(provider) = user_provider {
-        let provider = auth::user_provider_from_option(provider).context(IllegalAuthConfigSnafu)?;
-        plugins.insert::<UserProviderRef>(provider);
-    }
-    Ok(plugins)
-}
-
 #[cfg(test)]
 mod tests {
     use std::io::Write;
@@ -218,6 +217,7 @@ mod tests {
     use common_base::readable_size::ReadableSize;
     use common_test_util::temp_dir::create_named_temp_file;
     use frontend::service_config::GrpcOptions;
+    use servers::http::HttpOptions;
 
     use super::*;
     use crate::options::ENV_VAR_SEP;
@@ -303,14 +303,17 @@ mod tests {
 
     #[tokio::test]
     async fn test_try_from_start_command_to_anymap() {
-        let command = StartCommand {
+        let mut fe_opts = FrontendOptions {
+            http: HttpOptions {
+                disable_dashboard: false,
+                ..Default::default()
+            },
             user_provider: Some("static_user_provider:cmd:test=test".to_string()),
-            disable_dashboard: Some(false),
             ..Default::default()
         };
 
-        let plugins = load_frontend_plugins(&command.user_provider);
-        let plugins = plugins.unwrap();
+        let plugins = plugins::setup_frontend_plugins(&mut fe_opts).await.unwrap();
+
         let provider = plugins.get::<UserProviderRef>().unwrap();
         let result = provider
             .authenticate(
@@ -350,8 +353,8 @@ mod tests {
             addr = "127.0.0.1:4000"
 
             [meta_client]
-            timeout_millis = 3000
-            connect_timeout_millis = 5000
+            timeout = "3s"
+            connect_timeout = "5s"
             tcp_nodelay = true
 
             [mysql]

src/cmd/src/metasrv.rs

@@ -20,7 +20,7 @@ use meta_srv::bootstrap::MetaSrvInstance;
 use meta_srv::metasrv::MetaSrvOptions;
 use snafu::ResultExt;
 
-use crate::error::{self, Result};
+use crate::error::{self, Result, StartMetaServerSnafu};
 use crate::options::{Options, TopLevelOptions};
 
 pub struct Instance {
@@ -29,10 +29,10 @@ pub struct Instance {
 
 impl Instance {
     pub async fn start(&mut self) -> Result<()> {
-        self.instance
-            .start()
+        plugins::start_meta_srv_plugins(self.instance.plugins())
             .await
-            .context(error::StartMetaServerSnafu)
+            .context(StartMetaServerSnafu)?;
+        self.instance.start().await.context(StartMetaServerSnafu)
     }
 
     pub async fn stop(&self) -> Result<()> {
@@ -158,12 +158,15 @@ impl StartCommand {
         Ok(Options::Metasrv(Box::new(opts)))
     }
 
-    async fn build(self, opts: MetaSrvOptions) -> Result<Instance> {
-        logging::info!("MetaSrv start command: {:#?}", self);
+    async fn build(self, mut opts: MetaSrvOptions) -> Result<Instance> {
+        let plugins = plugins::setup_meta_srv_plugins(&mut opts)
+            .await
+            .context(StartMetaServerSnafu)?;
 
+        logging::info!("MetaSrv start command: {:#?}", self);
         logging::info!("MetaSrv options: {:#?}", opts);
 
-        let instance = MetaSrvInstance::new(opts)
+        let instance = MetaSrvInstance::new(opts, plugins)
             .await
             .context(error::BuildMetaServerSnafu)?;
 

src/cmd/src/options.rs

@@ -21,7 +21,7 @@ use meta_srv::metasrv::MetaSrvOptions;
 use serde::{Deserialize, Serialize};
 use snafu::ResultExt;
 
-use crate::error::{LoadLayeredConfigSnafu, Result};
+use crate::error::{LoadLayeredConfigSnafu, Result, SerdeJsonSnafu};
 
 pub const ENV_VAR_SEP: &str = "__";
 pub const ENV_LIST_SEP: &str = ",";
@@ -30,7 +30,7 @@ pub const ENV_LIST_SEP: &str = ",";
 pub struct MixOptions {
     pub data_home: String,
     pub procedure: ProcedureConfig,
-    pub kv_store: KvStoreConfig,
+    pub metadata_store: KvStoreConfig,
     pub frontend: FrontendOptions,
     pub datanode: DatanodeOptions,
     pub logging: LoggingOptions,
@@ -94,9 +94,16 @@ impl Options {
                 .ignore_empty(true)
         };
 
+        // Workaround: Replacement for `Config::try_from(&default_opts)` due to
+        // `ConfigSerializer` cannot handle the case of an empty struct contained
+        // within an iterative structure.
+        // See: https://github.com/mehcode/config-rs/issues/461
+        let json_str = serde_json::to_string(&default_opts).context(SerdeJsonSnafu)?;
+        let default_config = File::from_str(&json_str, FileFormat::Json);
+
         // Add default values and environment variables as the sources of the configuration.
         let mut layered_config = Config::builder()
-            .add_source(Config::try_from(&default_opts).context(LoadLayeredConfigSnafu)?)
+            .add_source(default_config)
             .add_source(env_source);
 
         // Add config file as the source of the configuration if it is specified.
@@ -137,8 +144,8 @@ mod tests {
             mysql_runtime_size = 2
 
             [meta_client]
-            timeout_millis = 3000
-            connect_timeout_millis = 5000
+            timeout = "3s"
+            connect_timeout = "5s"
             tcp_nodelay = true
 
             [wal]
@@ -256,6 +263,9 @@ mod tests {
                     ]
                 );
 
+                // Should be the values from config file, not environment variables.
+                assert_eq!(opts.wal.dir.unwrap(), "/tmp/greptimedb/wal");
+
                 // Should be default values.
                 assert_eq!(opts.node_id, None);
             },

src/cmd/src/standalone.rs

@@ -13,12 +13,13 @@
 // limitations under the License.
 
 use std::sync::Arc;
+use std::{fs, path};
 
 use catalog::kvbackend::KvBackendCatalogManager;
 use catalog::CatalogManagerRef;
 use clap::Parser;
 use common_base::Plugins;
-use common_config::{kv_store_dir, KvStoreConfig, WalConfig};
+use common_config::{metadata_store_dir, KvStoreConfig, WalConfig};
 use common_meta::cache_invalidator::DummyKvCacheInvalidator;
 use common_meta::kv_backend::KvBackendRef;
 use common_procedure::ProcedureManagerRef;
@@ -41,10 +42,10 @@ use servers::Mode;
 use snafu::ResultExt;
 
 use crate::error::{
-    IllegalConfigSnafu, InitMetadataSnafu, Result, ShutdownDatanodeSnafu, ShutdownFrontendSnafu,
-    StartDatanodeSnafu, StartFrontendSnafu,
+    CreateDirSnafu, IllegalConfigSnafu, InitMetadataSnafu, Result, ShutdownDatanodeSnafu,
+    ShutdownFrontendSnafu, StartDatanodeSnafu, StartFrontendSnafu, StartProcedureManagerSnafu,
+    StopProcedureManagerSnafu,
 };
-use crate::frontend::load_frontend_plugins;
 use crate::options::{MixOptions, Options, TopLevelOptions};
 
 #[derive(Parser)]
@@ -96,9 +97,10 @@ pub struct StandaloneOptions {
     pub prom_store: PromStoreOptions,
     pub wal: WalConfig,
     pub storage: StorageConfig,
-    pub kv_store: KvStoreConfig,
+    pub metadata_store: KvStoreConfig,
     pub procedure: ProcedureConfig,
     pub logging: LoggingOptions,
+    pub user_provider: Option<String>,
     /// Options for different store engines.
     pub region_engine: Vec<RegionEngineConfig>,
 }
@@ -117,9 +119,10 @@ impl Default for StandaloneOptions {
             prom_store: PromStoreOptions::default(),
             wal: WalConfig::default(),
             storage: StorageConfig::default(),
-            kv_store: KvStoreConfig::default(),
+            metadata_store: KvStoreConfig::default(),
             procedure: ProcedureConfig::default(),
             logging: LoggingOptions::default(),
+            user_provider: None,
             region_engine: vec![
                 RegionEngineConfig::Mito(MitoConfig::default()),
                 RegionEngineConfig::File(FileEngineConfig::default()),
@@ -141,6 +144,7 @@ impl StandaloneOptions {
             prom_store: self.prom_store,
             meta_client: None,
             logging: self.logging,
+            user_provider: self.user_provider,
             ..Default::default()
         }
     }
@@ -160,6 +164,7 @@ impl StandaloneOptions {
 pub struct Instance {
     datanode: Datanode,
     frontend: FeInstance,
+    procedure_manager: ProcedureManagerRef,
 }
 
 impl Instance {
@@ -168,6 +173,11 @@ impl Instance {
         self.datanode.start().await.context(StartDatanodeSnafu)?;
         info!("Datanode instance started");
 
+        self.procedure_manager
+            .start()
+            .await
+            .context(StartProcedureManagerSnafu)?;
+
         self.frontend.start().await.context(StartFrontendSnafu)?;
         Ok(())
     }
@@ -178,6 +188,11 @@ impl Instance {
             .await
             .context(ShutdownFrontendSnafu)?;
 
+        self.procedure_manager
+            .stop()
+            .await
+            .context(StopProcedureManagerSnafu)?;
+
         self.datanode
             .shutdown()
             .await
@@ -278,7 +293,10 @@ impl StartCommand {
         if self.influxdb_enable {
             opts.influxdb.enable = self.influxdb_enable;
         }
-        let kv_store = opts.kv_store.clone();
+
+        opts.user_provider = self.user_provider.clone();
+
+        let metadata_store = opts.metadata_store.clone();
         let procedure = opts.procedure.clone();
         let frontend = opts.clone().frontend_options();
         let logging = opts.logging.clone();
@@ -286,7 +304,7 @@ impl StartCommand {
 
         Ok(Options::Standalone(Box::new(MixOptions {
             procedure,
-            kv_store,
+            metadata_store,
             data_home: datanode.storage.data_home.to_string(),
             frontend,
             datanode,
@@ -298,8 +316,11 @@ impl StartCommand {
     #[allow(unused_variables)]
     #[allow(clippy::diverging_sub_expression)]
     async fn build(self, opts: MixOptions) -> Result<Instance> {
-        let plugins = Arc::new(load_frontend_plugins(&self.user_provider)?);
-        let fe_opts = opts.frontend;
+        let mut fe_opts = opts.frontend;
+        let fe_plugins = plugins::setup_frontend_plugins(&mut fe_opts)
+            .await
+            .context(StartFrontendSnafu)?;
+
         let dn_opts = opts.datanode;
 
         info!("Standalone start command: {:#?}", self);
@@ -308,14 +329,22 @@ impl StartCommand {
             fe_opts, dn_opts
         );
 
-        let kv_dir = kv_store_dir(&opts.data_home);
-        let (kv_store, procedure_manager) =
-            FeInstance::try_build_standalone_components(kv_dir, opts.kv_store, opts.procedure)
-                .await
-                .context(StartFrontendSnafu)?;
+        // Ensure the data_home directory exists.
+        fs::create_dir_all(path::Path::new(&opts.data_home)).context(CreateDirSnafu {
+            dir: &opts.data_home,
+        })?;
+
+        let metadata_dir = metadata_store_dir(&opts.data_home);
+        let (kv_store, procedure_manager) = FeInstance::try_build_standalone_components(
+            metadata_dir,
+            opts.metadata_store,
+            opts.procedure,
+        )
+        .await
+        .context(StartFrontendSnafu)?;
 
         let datanode =
-            DatanodeBuilder::new(dn_opts.clone(), Some(kv_store.clone()), plugins.clone())
+            DatanodeBuilder::new(dn_opts.clone(), Some(kv_store.clone()), Default::default())
                 .build()
                 .await
                 .context(StartDatanodeSnafu)?;
@@ -335,9 +364,9 @@ impl StartCommand {
 
         // TODO: build frontend instance like in distributed mode
         let mut frontend = build_frontend(
-            plugins,
+            fe_plugins,
             kv_store,
-            procedure_manager,
+            procedure_manager.clone(),
             catalog_manager,
             region_server,
         )
@@ -348,13 +377,17 @@ impl StartCommand {
             .await
             .context(StartFrontendSnafu)?;
 
-        Ok(Instance { datanode, frontend })
+        Ok(Instance {
+            datanode,
+            frontend,
+            procedure_manager,
+        })
     }
 }
 
 /// Build frontend instance in standalone mode
 async fn build_frontend(
-    plugins: Arc<Plugins>,
+    plugins: Plugins,
     kv_store: KvBackendRef,
     procedure_manager: ProcedureManagerRef,
     catalog_manager: CatalogManagerRef,
@@ -388,13 +421,13 @@ mod tests {
 
     #[tokio::test]
     async fn test_try_from_start_command_to_anymap() {
-        let command = StartCommand {
+        let mut fe_opts = FrontendOptions {
             user_provider: Some("static_user_provider:cmd:test=test".to_string()),
             ..Default::default()
         };
 
-        let plugins = load_frontend_plugins(&command.user_provider);
-        let plugins = plugins.unwrap();
+        let plugins = plugins::setup_frontend_plugins(&mut fe_opts).await.unwrap();
+
         let provider = plugins.get::<UserProviderRef>().unwrap();
         let result = provider
             .authenticate(
@@ -476,6 +509,8 @@ mod tests {
         assert_eq!(None, fe_opts.mysql.reject_no_database);
         assert!(fe_opts.influxdb.enable);
 
+        assert_eq!("/tmp/greptimedb/test/wal", dn_opts.wal.dir.unwrap());
+
         match &dn_opts.storage.store {
             datanode::config::ObjectStoreConfig::S3(s3_config) => {
                 assert_eq!(
@@ -593,4 +628,25 @@ mod tests {
             },
         );
     }
+
+    #[test]
+    fn test_load_default_standalone_options() {
+        let options: StandaloneOptions =
+            Options::load_layered_options(None, "GREPTIMEDB_FRONTEND", None).unwrap();
+        let default_options = StandaloneOptions::default();
+        assert_eq!(options.mode, default_options.mode);
+        assert_eq!(options.enable_telemetry, default_options.enable_telemetry);
+        assert_eq!(options.http, default_options.http);
+        assert_eq!(options.grpc, default_options.grpc);
+        assert_eq!(options.mysql, default_options.mysql);
+        assert_eq!(options.postgres, default_options.postgres);
+        assert_eq!(options.opentsdb, default_options.opentsdb);
+        assert_eq!(options.influxdb, default_options.influxdb);
+        assert_eq!(options.prom_store, default_options.prom_store);
+        assert_eq!(options.wal, default_options.wal);
+        assert_eq!(options.metadata_store, default_options.metadata_store);
+        assert_eq!(options.procedure, default_options.procedure);
+        assert_eq!(options.logging, default_options.logging);
+        assert_eq!(options.region_engine, default_options.region_engine);
+    }
 }

src/common/base/Cargo.toml

@@ -9,6 +9,7 @@ anymap = "1.0.0-beta.2"
 bitvec = "1.0"
 bytes = { version = "1.1", features = ["serde"] }
 common-error = { workspace = true }
+common-macro = { workspace = true }
 paste = "1.0"
 serde = { version = "1.0", features = ["derive"] }
 snafu.workspace = true

src/common/base/src/buffer.rs

@@ -17,11 +17,13 @@ use std::io::{Read, Write};
 
 use bytes::{Buf, BufMut, BytesMut};
 use common_error::ext::ErrorExt;
+use common_macro::stack_trace_debug;
 use paste::paste;
 use snafu::{ensure, Location, ResultExt, Snafu};
 
-#[derive(Debug, Snafu)]
+#[derive(Snafu)]
 #[snafu(visibility(pub))]
+#[stack_trace_debug]
 pub enum Error {
     #[snafu(display(
         "Destination buffer overflow, src_len: {}, dst_len: {}",
@@ -39,7 +41,8 @@ pub enum Error {
 
     #[snafu(display("IO operation reach EOF"))]
     Eof {
-        source: std::io::Error,
+        #[snafu(source)]
+        error: std::io::Error,
         location: Location,
     },
 }

src/common/base/src/lib.rs

@@ -23,6 +23,8 @@ use std::sync::{Arc, Mutex, MutexGuard};
 
 pub use bit_vec::BitVec;
 
+/// [`Plugins`] is a wrapper of Arc contents.
+/// Make it Cloneable and we can treat it like an Arc struct.
 #[derive(Default, Clone)]
 pub struct Plugins {
     inner: Arc<Mutex<anymap::Map<dyn Any + Send + Sync>>>,

src/common/base/src/readable_size.rs

@@ -14,10 +14,9 @@
 // See the License for the specific language governing permissions and
 // limitations under the License.
 
-// This file is copied from https://github.com/tikv/raft-engine/blob/8dd2a39f359ff16f5295f35343f626e0c10132fa/src/util.rs without any modification.
+// This file is copied from https://github.com/tikv/raft-engine/blob/8dd2a39f359ff16f5295f35343f626e0c10132fa/src/util.rs
 
-use std::fmt;
-use std::fmt::{Display, Write};
+use std::fmt::{self, Debug, Display, Write};
 use std::ops::{Div, Mul};
 use std::str::FromStr;
 
@@ -34,7 +33,7 @@ pub const GIB: u64 = MIB * BINARY_DATA_MAGNITUDE;
 pub const TIB: u64 = GIB * BINARY_DATA_MAGNITUDE;
 pub const PIB: u64 = TIB * BINARY_DATA_MAGNITUDE;
 
-#[derive(Clone, Debug, Copy, PartialEq, Eq, PartialOrd)]
+#[derive(Clone, Copy, PartialEq, Eq, PartialOrd)]
 pub struct ReadableSize(pub u64);
 
 impl ReadableSize {
@@ -155,6 +154,12 @@ impl FromStr for ReadableSize {
     }
 }
 
+impl Debug for ReadableSize {
+    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
+        write!(f, "{}", self)
+    }
+}
+
 impl Display for ReadableSize {
     fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
         if self.0 >= PIB {

src/common/catalog/Cargo.toml

@@ -6,6 +6,7 @@ license.workspace = true
 
 [dependencies]
 common-error = { workspace = true }
+common-macro = { workspace = true }
 serde.workspace = true
 serde_json = "1.0"
 snafu = { version = "0.7", features = ["backtraces"] }

src/common/catalog/src/error.rs

@@ -16,10 +16,12 @@ use std::any::Any;
 
 use common_error::ext::ErrorExt;
 use common_error::status_code::StatusCode;
+use common_macro::stack_trace_debug;
 use snafu::{Location, Snafu};
 
-#[derive(Debug, Snafu)]
+#[derive(Snafu)]
 #[snafu(visibility(pub))]
+#[stack_trace_debug]
 pub enum Error {
     #[snafu(display("Invalid full table name: {}", table_name))]
     InvalidFullTableName {

src/common/catalog/src/lib.rs

@@ -13,9 +13,6 @@
 // limitations under the License.
 
 use consts::DEFAULT_CATALOG_NAME;
-use snafu::ensure;
-
-use crate::error::Result;
 
 pub mod consts;
 pub mod error;
@@ -26,17 +23,6 @@ pub fn format_full_table_name(catalog: &str, schema: &str, table: &str) -> Strin
     format!("{catalog}.{schema}.{table}")
 }
 
-pub fn parse_full_table_name(table_name: &str) -> Result<(&str, &str, &str)> {
-    let result = table_name.split('.').collect::<Vec<_>>();
-
-    ensure!(
-        result.len() == 3,
-        error::InvalidFullTableNameSnafu { table_name }
-    );
-
-    Ok((result[0], result[1], result[2]))
-}
-
 /// Build db name from catalog and schema string
 pub fn build_db_string(catalog: &str, schema: &str) -> String {
     if catalog == DEFAULT_CATALOG_NAME {

src/common/config/src/lib.rs

@@ -17,9 +17,11 @@ use std::time::Duration;
 use common_base::readable_size::ReadableSize;
 use serde::{Deserialize, Serialize};
 
-#[derive(Debug, Clone, Serialize, Deserialize)]
+#[derive(Debug, Clone, Serialize, Deserialize, PartialEq, Eq)]
 #[serde(default)]
 pub struct WalConfig {
+    // wal directory
+    pub dir: Option<String>,
     // wal file size in bytes
     pub file_size: ReadableSize,
     // wal purge threshold in bytes
@@ -36,7 +38,8 @@ pub struct WalConfig {
 impl Default for WalConfig {
     fn default() -> Self {
         Self {
-            file_size: ReadableSize::mb(256),     // log file size 256MB
+            dir: None,
+            file_size: ReadableSize::mb(256), // log file size 256MB
             purge_threshold: ReadableSize::gb(4), // purge threshold 4GB
             purge_interval: Duration::from_secs(600),
             read_batch_size: 128,
@@ -45,11 +48,11 @@ impl Default for WalConfig {
     }
 }
 
-pub fn kv_store_dir(store_dir: &str) -> String {
-    format!("{store_dir}/kv")
+pub fn metadata_store_dir(store_dir: &str) -> String {
+    format!("{store_dir}/metadata")
 }
 
-#[derive(Debug, Clone, Serialize, Deserialize)]
+#[derive(Debug, Clone, Serialize, Deserialize, PartialEq, Eq)]
 #[serde(default)]
 pub struct KvStoreConfig {
     // Kv file size in bytes

src/common/datasource/Cargo.toml

@@ -18,6 +18,7 @@ async-compression = { version = "0.3", features = [
 async-trait.workspace = true
 bytes = "1.1"
 common-error = { workspace = true }
+common-macro = { workspace = true }
 common-runtime = { workspace = true }
 datafusion.workspace = true
 derive_builder.workspace = true

src/common/datasource/src/error.rs

@@ -17,12 +17,14 @@ use std::any::Any;
 use arrow_schema::ArrowError;
 use common_error::ext::ErrorExt;
 use common_error::status_code::StatusCode;
+use common_macro::stack_trace_debug;
 use datafusion::parquet::errors::ParquetError;
 use snafu::{Location, Snafu};
 use url::ParseError;
 
-#[derive(Debug, Snafu)]
+#[derive(Snafu)]
 #[snafu(visibility(pub))]
+#[stack_trace_debug]
 pub enum Error {
     #[snafu(display("Unsupported compression type: {}", compression_type))]
     UnsupportedCompressionType {
@@ -46,83 +48,96 @@ pub enum Error {
     #[snafu(display("Invalid url: {}", url))]
     InvalidUrl {
         url: String,
-        source: ParseError,
+        #[snafu(source)]
+        error: ParseError,
         location: Location,
     },
 
     #[snafu(display("Failed to build backend"))]
     BuildBackend {
-        source: object_store::Error,
+        #[snafu(source)]
+        error: object_store::Error,
         location: Location,
     },
 
     #[snafu(display("Failed to build orc reader"))]
     OrcReader {
         location: Location,
-        source: orc_rust::error::Error,
+        #[snafu(source)]
+        error: orc_rust::error::Error,
     },
 
     #[snafu(display("Failed to read object from path: {}", path))]
     ReadObject {
         path: String,
         location: Location,
-        source: object_store::Error,
+        #[snafu(source)]
+        error: object_store::Error,
     },
 
     #[snafu(display("Failed to write object to path: {}", path))]
     WriteObject {
         path: String,
         location: Location,
-        source: object_store::Error,
+        #[snafu(source)]
+        error: object_store::Error,
     },
 
     #[snafu(display("Failed to write"))]
     AsyncWrite {
-        source: std::io::Error,
+        #[snafu(source)]
+        error: std::io::Error,
         location: Location,
     },
 
     #[snafu(display("Failed to write record batch"))]
     WriteRecordBatch {
         location: Location,
-        source: ArrowError,
+        #[snafu(source)]
+        error: ArrowError,
     },
 
     #[snafu(display("Failed to encode record batch"))]
     EncodeRecordBatch {
         location: Location,
-        source: ParquetError,
+        #[snafu(source)]
+        error: ParquetError,
     },
 
     #[snafu(display("Failed to read record batch"))]
     ReadRecordBatch {
         location: Location,
-        source: datafusion::error::DataFusionError,
+        #[snafu(source)]
+        error: datafusion::error::DataFusionError,
     },
 
     #[snafu(display("Failed to read parquet"))]
     ReadParquetSnafu {
         location: Location,
-        source: datafusion::parquet::errors::ParquetError,
+        #[snafu(source)]
+        error: datafusion::parquet::errors::ParquetError,
     },
 
     #[snafu(display("Failed to convert parquet to schema"))]
     ParquetToSchema {
         location: Location,
-        source: datafusion::parquet::errors::ParquetError,
+        #[snafu(source)]
+        error: datafusion::parquet::errors::ParquetError,
     },
 
     #[snafu(display("Failed to infer schema from file"))]
     InferSchema {
         location: Location,
-        source: arrow_schema::ArrowError,
+        #[snafu(source)]
+        error: arrow_schema::ArrowError,
     },
 
     #[snafu(display("Failed to list object in path: {}", path))]
     ListObjects {
         path: String,
         location: Location,
-        source: object_store::Error,
+        #[snafu(source)]
+        error: object_store::Error,
     },
 
     #[snafu(display("Invalid connection: {}", msg))]
@@ -131,7 +146,8 @@ pub enum Error {
     #[snafu(display("Failed to join handle"))]
     JoinHandle {
         location: Location,
-        source: tokio::task::JoinError,
+        #[snafu(source)]
+        error: tokio::task::JoinError,
     },
 
     #[snafu(display("Failed to parse format {} with value: {}", key, value))]
@@ -143,7 +159,8 @@ pub enum Error {
 
     #[snafu(display("Failed to merge schema"))]
     MergeSchema {
-        source: arrow_schema::ArrowError,
+        #[snafu(source)]
+        error: arrow_schema::ArrowError,
         location: Location,
     },
 

src/common/error/src/ext.rs

@@ -13,11 +13,12 @@
 // limitations under the License.
 
 use std::any::Any;
+use std::sync::Arc;
 
 use crate::status_code::StatusCode;
 
 /// Extension to [`Error`](std::error::Error) in std.
-pub trait ErrorExt: std::error::Error {
+pub trait ErrorExt: StackError {
     /// Map this error to [StatusCode].
     fn status_code(&self) -> StatusCode {
         StatusCode::Unknown
@@ -33,6 +34,71 @@ pub trait ErrorExt: std::error::Error {
     /// Returns the error as [Any](std::any::Any) so that it can be
     /// downcast to a specific implementation.
     fn as_any(&self) -> &dyn Any;
+
+    fn output_msg(&self) -> String
+    where
+        Self: Sized,
+    {
+        match self.status_code() {
+            StatusCode::Unknown | StatusCode::Unexpected | StatusCode::Internal => {
+                // masks internal error from end user
+                format!("Internal error: {}", self.status_code() as u32)
+            }
+            _ => {
+                let error = self.last();
+                if let Some(external_error) = error.source() {
+                    let external_root = external_error.sources().last().unwrap();
+
+                    if error.to_string().is_empty() {
+                        format!("{external_root}")
+                    } else {
+                        format!("{error}: {external_root}")
+                    }
+                } else {
+                    format!("{error}")
+                }
+            }
+        }
+    }
+}
+
+pub trait StackError: std::error::Error {
+    fn debug_fmt(&self, layer: usize, buf: &mut Vec<String>);
+
+    fn next(&self) -> Option<&dyn StackError>;
+
+    fn last(&self) -> &dyn StackError
+    where
+        Self: Sized,
+    {
+        let Some(mut result) = self.next() else {
+            return self;
+        };
+        while let Some(err) = result.next() {
+            result = err;
+        }
+        result
+    }
+}
+
+impl<T: ?Sized + StackError> StackError for Arc<T> {
+    fn debug_fmt(&self, layer: usize, buf: &mut Vec<String>) {
+        self.as_ref().debug_fmt(layer, buf)
+    }
+
+    fn next(&self) -> Option<&dyn StackError> {
+        self.as_ref().next()
+    }
+}
+
+impl<T: StackError> StackError for Box<T> {
+    fn debug_fmt(&self, layer: usize, buf: &mut Vec<String>) {
+        self.as_ref().debug_fmt(layer, buf)
+    }
+
+    fn next(&self) -> Option<&dyn StackError> {
+        self.as_ref().next()
+    }
 }
 
 /// An opaque boxed error based on errors that implement [ErrorExt] trait.
@@ -90,6 +156,16 @@ impl crate::snafu::ErrorCompat for BoxedError {
     }
 }
 
+impl StackError for BoxedError {
+    fn debug_fmt(&self, layer: usize, buf: &mut Vec<String>) {
+        self.inner.debug_fmt(layer, buf)
+    }
+
+    fn next(&self) -> Option<&dyn StackError> {
+        self.inner.next()
+    }
+}
+
 /// Error type with plain error message
 #[derive(Debug)]
 pub struct PlainError {
@@ -128,3 +204,13 @@ impl crate::ext::ErrorExt for PlainError {
         self as _
     }
 }
+
+impl StackError for PlainError {
+    fn debug_fmt(&self, layer: usize, buf: &mut Vec<String>) {
+        buf.push(format!("{}: {}", layer, self.msg))
+    }
+
+    fn next(&self) -> Option<&dyn StackError> {
+        None
+    }
+}

src/common/error/src/format.rs

@@ -50,6 +50,7 @@ mod tests {
     use snafu::{GenerateImplicitData, Location};
 
     use super::*;
+    use crate::ext::StackError;
 
     #[derive(Debug, Snafu)]
     #[snafu(display("This is a leaf error"))]
@@ -65,6 +66,14 @@ mod tests {
         }
     }
 
+    impl StackError for Leaf {
+        fn debug_fmt(&self, _: usize, _: &mut Vec<String>) {}
+
+        fn next(&self) -> Option<&dyn StackError> {
+            None
+        }
+    }
+
     #[derive(Debug, Snafu)]
     #[snafu(display("This is a leaf with location"))]
     struct LeafWithLocation {
@@ -81,6 +90,14 @@ mod tests {
         }
     }
 
+    impl StackError for LeafWithLocation {
+        fn debug_fmt(&self, _: usize, _: &mut Vec<String>) {}
+
+        fn next(&self) -> Option<&dyn StackError> {
+            None
+        }
+    }
+
     #[derive(Debug, Snafu)]
     #[snafu(display("Internal error"))]
     struct Internal {
@@ -99,6 +116,17 @@ mod tests {
         }
     }
 
+    impl StackError for Internal {
+        fn debug_fmt(&self, layer: usize, buf: &mut Vec<String>) {
+            buf.push(format!("{}: Internal error, at {}", layer, self.location));
+            self.source.debug_fmt(layer + 1, buf);
+        }
+
+        fn next(&self) -> Option<&dyn StackError> {
+            Some(&self.source)
+        }
+    }
+
     #[test]
     fn test_debug_format() {
         let err = Leaf;

src/common/error/src/lib.rs

@@ -11,6 +11,7 @@
 // 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.
+#![feature(error_iter)]
 
 pub mod ext;
 pub mod format;

src/common/error/src/mock.rs

@@ -19,7 +19,7 @@ use std::fmt;
 
 use snafu::Location;
 
-use crate::ext::ErrorExt;
+use crate::ext::{ErrorExt, StackError};
 use crate::status_code::StatusCode;
 
 /// A mock error mainly for test.
@@ -69,3 +69,11 @@ impl ErrorExt for MockError {
         self
     }
 }
+
+impl StackError for MockError {
+    fn debug_fmt(&self, _: usize, _: &mut Vec<String>) {}
+
+    fn next(&self) -> Option<&dyn StackError> {
+        None
+    }
+}

src/common/function/src/scalars/math.rs

@@ -58,9 +58,15 @@ impl Function for RangeFunction {
         "range_fn"
     }
 
-    // range_fn will never been used, return_type could be arbitrary value, is not important
-    fn return_type(&self, _input_types: &[ConcreteDataType]) -> Result<ConcreteDataType> {
-        Ok(ConcreteDataType::float64_datatype())
+    // The first argument to range_fn is the expression to be evaluated
+    fn return_type(&self, input_types: &[ConcreteDataType]) -> Result<ConcreteDataType> {
+        input_types
+            .first()
+            .cloned()
+            .ok_or(DataFusionError::Internal(
+                "No expr found in range_fn".into(),
+            ))
+            .context(GeneralDataFusionSnafu)
     }
 
     /// `range_fn` will never been used. As long as a legal signature is returned, the specific content of the signature does not matter.

src/common/function/src/scalars/timestamp.rs

@@ -12,8 +12,10 @@
 // See the License for the specific language governing permissions and
 // limitations under the License.
 use std::sync::Arc;
+mod greatest;
 mod to_unixtime;
 
+use greatest::GreatestFunction;
 use to_unixtime::ToUnixtimeFunction;
 
 use crate::scalars::function_registry::FunctionRegistry;
@@ -23,5 +25,6 @@ pub(crate) struct TimestampFunction;
 impl TimestampFunction {
     pub fn register(registry: &FunctionRegistry) {
         registry.register(Arc::new(ToUnixtimeFunction));
+        registry.register(Arc::new(GreatestFunction));
     }
 }

src/common/function/src/scalars/timestamp/greatest.rs

@@ -0,0 +1,175 @@
+// 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::fmt::{self};
+
+use common_query::error::{
+    self, ArrowComputeSnafu, InvalidFuncArgsSnafu, Result, UnsupportedInputDataTypeSnafu,
+};
+use common_query::prelude::{Signature, Volatility};
+use datatypes::arrow::array::AsArray;
+use datatypes::arrow::compute::cast;
+use datatypes::arrow::compute::kernels::comparison::gt_dyn;
+use datatypes::arrow::compute::kernels::zip;
+use datatypes::arrow::datatypes::{DataType as ArrowDataType, Date32Type};
+use datatypes::prelude::ConcreteDataType;
+use datatypes::vectors::{Helper, VectorRef};
+use snafu::{ensure, ResultExt};
+
+use crate::scalars::function::{Function, FunctionContext};
+
+#[derive(Clone, Debug, Default)]
+pub struct GreatestFunction;
+
+const NAME: &str = "greatest";
+
+impl Function for GreatestFunction {
+    fn name(&self) -> &str {
+        NAME
+    }
+
+    fn return_type(&self, _input_types: &[ConcreteDataType]) -> Result<ConcreteDataType> {
+        Ok(ConcreteDataType::date_datatype())
+    }
+
+    fn signature(&self) -> Signature {
+        Signature::uniform(
+            2,
+            vec![
+                ConcreteDataType::string_datatype(),
+                ConcreteDataType::date_datatype(),
+            ],
+            Volatility::Immutable,
+        )
+    }
+
+    fn eval(&self, _func_ctx: FunctionContext, columns: &[VectorRef]) -> Result<VectorRef> {
+        ensure!(
+            columns.len() == 2,
+            InvalidFuncArgsSnafu {
+                err_msg: format!(
+                    "The length of the args is not correct, expect exactly two, have: {}",
+                    columns.len()
+                ),
+            }
+        );
+        match columns[0].data_type() {
+            ConcreteDataType::String(_) => {
+                let column1 = cast(&columns[0].to_arrow_array(), &ArrowDataType::Date32)
+                    .context(ArrowComputeSnafu)?;
+                let column1 = column1.as_primitive::<Date32Type>();
+                let column2 = cast(&columns[1].to_arrow_array(), &ArrowDataType::Date32)
+                    .context(ArrowComputeSnafu)?;
+                let column2 = column2.as_primitive::<Date32Type>();
+                let boolean_array = gt_dyn(&column1, &column2).context(ArrowComputeSnafu)?;
+                let result =
+                    zip::zip(&boolean_array, &column1, &column2).context(ArrowComputeSnafu)?;
+                Ok(Helper::try_into_vector(&result).context(error::FromArrowArraySnafu)?)
+            }
+            ConcreteDataType::Date(_) => {
+                let column1 = columns[0].to_arrow_array();
+                let column1 = column1.as_primitive::<Date32Type>();
+                let column2 = columns[1].to_arrow_array();
+                let column2 = column2.as_primitive::<Date32Type>();
+                let boolean_array = gt_dyn(&column1, &column2).context(ArrowComputeSnafu)?;
+                let result =
+                    zip::zip(&boolean_array, &column1, &column2).context(ArrowComputeSnafu)?;
+                Ok(Helper::try_into_vector(&result).context(error::FromArrowArraySnafu)?)
+            }
+            _ => UnsupportedInputDataTypeSnafu {
+                function: NAME,
+                datatypes: columns.iter().map(|c| c.data_type()).collect::<Vec<_>>(),
+            }
+            .fail(),
+        }
+    }
+}
+
+impl fmt::Display for GreatestFunction {
+    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
+        write!(f, "GREATEST")
+    }
+}
+
+#[cfg(test)]
+mod tests {
+    use std::str::FromStr;
+    use std::sync::Arc;
+
+    use common_time::Date;
+    use datatypes::prelude::ConcreteDataType;
+    use datatypes::types::DateType;
+    use datatypes::value::Value;
+    use datatypes::vectors::{DateVector, StringVector, Vector};
+
+    use super::GreatestFunction;
+    use crate::scalars::function::FunctionContext;
+    use crate::scalars::Function;
+
+    #[test]
+    fn test_greatest_takes_string_vector() {
+        let function = GreatestFunction;
+        assert_eq!(
+            function.return_type(&[]).unwrap(),
+            ConcreteDataType::Date(DateType)
+        );
+        let columns = vec![
+            Arc::new(StringVector::from(vec![
+                "1970-01-01".to_string(),
+                "2012-12-23".to_string(),
+            ])) as _,
+            Arc::new(StringVector::from(vec![
+                "2001-02-01".to_string(),
+                "1999-01-01".to_string(),
+            ])) as _,
+        ];
+
+        let result = function.eval(FunctionContext::default(), &columns).unwrap();
+        let result = result.as_any().downcast_ref::<DateVector>().unwrap();
+        assert_eq!(result.len(), 2);
+        assert_eq!(
+            result.get(0),
+            Value::Date(Date::from_str("2001-02-01").unwrap())
+        );
+        assert_eq!(
+            result.get(1),
+            Value::Date(Date::from_str("2012-12-23").unwrap())
+        );
+    }
+
+    #[test]
+    fn test_greatest_takes_date_vector() {
+        let function = GreatestFunction;
+        assert_eq!(
+            function.return_type(&[]).unwrap(),
+            ConcreteDataType::Date(DateType)
+        );
+        let columns = vec![
+            Arc::new(DateVector::from_slice(vec![-1, 2])) as _,
+            Arc::new(DateVector::from_slice(vec![0, 1])) as _,
+        ];
+
+        let result = function.eval(FunctionContext::default(), &columns).unwrap();
+        let result = result.as_any().downcast_ref::<DateVector>().unwrap();
+        assert_eq!(result.len(), 2);
+        assert_eq!(
+            result.get(0),
+            Value::Date(Date::from_str("1970-01-01").unwrap())
+        );
+        assert_eq!(
+            result.get(1),
+            Value::Date(Date::from_str("1970-01-03").unwrap())
+        );
+    }
+}

src/common/greptimedb-telemetry/src/lib.rs

@@ -15,6 +15,8 @@
 use std::env;
 use std::io::ErrorKind;
 use std::path::{Path, PathBuf};
+use std::sync::atomic::{AtomicBool, Ordering};
+use std::sync::Arc;
 use std::time::Duration;
 
 use common_runtime::error::{Error, Result};
@@ -24,7 +26,7 @@ use reqwest::{Client, Response};
 use serde::{Deserialize, Serialize};
 
 /// The URL to report telemetry data.
-pub const TELEMETRY_URL: &str = "https://api.greptime.cloud/db/otel/statistics";
+pub const TELEMETRY_URL: &str = "https://telemetry.greptimestats.com/db/otel/statistics";
 /// The local installation uuid cache file
 const UUID_FILE_NAME: &str = ".greptimedb-telemetry-uuid";
 
@@ -36,13 +38,29 @@ const GREPTIMEDB_TELEMETRY_CLIENT_CONNECT_TIMEOUT: Duration = Duration::from_sec
 const GREPTIMEDB_TELEMETRY_CLIENT_REQUEST_TIMEOUT: Duration = Duration::from_secs(10);
 
 pub enum GreptimeDBTelemetryTask {
-    Enable(RepeatedTask<Error>),
+    Enable((RepeatedTask<Error>, Arc<AtomicBool>)),
     Disable,
 }
 
 impl GreptimeDBTelemetryTask {
-    pub fn enable(interval: Duration, task_fn: BoxedTaskFunction<Error>) -> Self {
-        GreptimeDBTelemetryTask::Enable(RepeatedTask::new(interval, task_fn))
+    pub fn should_report(&self, value: bool) {
+        match self {
+            GreptimeDBTelemetryTask::Enable((_, should_report)) => {
+                should_report.store(value, Ordering::Relaxed);
+            }
+            GreptimeDBTelemetryTask::Disable => {}
+        }
+    }
+
+    pub fn enable(
+        interval: Duration,
+        task_fn: BoxedTaskFunction<Error>,
+        should_report: Arc<AtomicBool>,
+    ) -> Self {
+        GreptimeDBTelemetryTask::Enable((
+            RepeatedTask::new(interval, task_fn).with_initial_delay(Some(Duration::ZERO)),
+            should_report,
+        ))
     }
 
     pub fn disable() -> Self {
@@ -51,7 +69,7 @@ impl GreptimeDBTelemetryTask {
 
     pub fn start(&self) -> Result<()> {
         match self {
-            GreptimeDBTelemetryTask::Enable(task) => {
+            GreptimeDBTelemetryTask::Enable((task, _)) => {
                 print_anonymous_usage_data_disclaimer();
                 task.start(common_runtime::bg_runtime())
             }
@@ -61,7 +79,7 @@ impl GreptimeDBTelemetryTask {
 
     pub async fn stop(&self) -> Result<()> {
         match self {
-            GreptimeDBTelemetryTask::Enable(task) => task.stop().await,
+            GreptimeDBTelemetryTask::Enable((task, _)) => task.stop().await,
             GreptimeDBTelemetryTask::Disable => Ok(()),
         }
     }
@@ -191,6 +209,7 @@ pub struct GreptimeDBTelemetry {
     client: Option<Client>,
     working_home: Option<String>,
     telemetry_url: &'static str,
+    should_report: Arc<AtomicBool>,
 }
 
 #[async_trait::async_trait]
@@ -200,13 +219,19 @@ impl TaskFunction<Error> for GreptimeDBTelemetry {
     }
 
     async fn call(&mut self) -> Result<()> {
-        self.report_telemetry_info().await;
+        if self.should_report.load(Ordering::Relaxed) {
+            self.report_telemetry_info().await;
+        }
         Ok(())
     }
 }
 
 impl GreptimeDBTelemetry {
-    pub fn new(working_home: Option<String>, statistics: Box<dyn Collector + Send + Sync>) -> Self {
+    pub fn new(
+        working_home: Option<String>,
+        statistics: Box<dyn Collector + Send + Sync>,
+        should_report: Arc<AtomicBool>,
+    ) -> Self {
         let client = Client::builder()
             .connect_timeout(GREPTIMEDB_TELEMETRY_CLIENT_CONNECT_TIMEOUT)
             .timeout(GREPTIMEDB_TELEMETRY_CLIENT_REQUEST_TIMEOUT)
@@ -216,6 +241,7 @@ impl GreptimeDBTelemetry {
             statistics,
             client: client.ok(),
             telemetry_url: TELEMETRY_URL,
+            should_report,
         }
     }
 
@@ -250,7 +276,8 @@ impl GreptimeDBTelemetry {
 mod tests {
     use std::convert::Infallible;
     use std::env;
-    use std::sync::atomic::AtomicUsize;
+    use std::sync::atomic::{AtomicBool, AtomicUsize};
+    use std::sync::Arc;
     use std::time::Duration;
 
     use common_test_util::ports;
@@ -370,7 +397,11 @@ mod tests {
         let working_home = working_home_temp.path().to_str().unwrap().to_string();
 
         let test_statistic = Box::new(TestStatistic);
-        let mut test_report = GreptimeDBTelemetry::new(Some(working_home.clone()), test_statistic);
+        let mut test_report = GreptimeDBTelemetry::new(
+            Some(working_home.clone()),
+            test_statistic,
+            Arc::new(AtomicBool::new(true)),
+        );
         let url = Box::leak(format!("{}:{}", "http://localhost", port).into_boxed_str());
         test_report.telemetry_url = url;
         let response = test_report.report_telemetry_info().await.unwrap();
@@ -384,7 +415,11 @@ mod tests {
         assert_eq!(1, body.nodes.unwrap());
 
         let failed_statistic = Box::new(FailedStatistic);
-        let mut failed_report = GreptimeDBTelemetry::new(Some(working_home), failed_statistic);
+        let mut failed_report = GreptimeDBTelemetry::new(
+            Some(working_home),
+            failed_statistic,
+            Arc::new(AtomicBool::new(true)),
+        );
         failed_report.telemetry_url = url;
         let response = failed_report.report_telemetry_info().await;
         assert!(response.is_none());

src/common/grpc-expr/Cargo.toml

@@ -10,6 +10,7 @@ async-trait.workspace = true
 common-base = { workspace = true }
 common-catalog = { workspace = true }
 common-error = { workspace = true }
+common-macro = { workspace = true }
 common-query = { workspace = true }
 common-telemetry = { workspace = true }
 common-time = { workspace = true }

src/common/grpc-expr/src/error.rs

@@ -16,10 +16,12 @@ use std::any::Any;
 
 use common_error::ext::ErrorExt;
 use common_error::status_code::StatusCode;
+use common_macro::stack_trace_debug;
 use snafu::{Location, Snafu};
 
-#[derive(Debug, Snafu)]
+#[derive(Snafu)]
 #[snafu(visibility(pub))]
+#[stack_trace_debug]
 pub enum Error {
     #[snafu(display("Illegal delete request, reason: {reason}"))]
     IllegalDeleteRequest { reason: String, location: Location },

src/common/grpc/Cargo.toml

@@ -11,6 +11,7 @@ async-trait = "0.1"
 backtrace = "0.3"
 common-base = { workspace = true }
 common-error = { workspace = true }
+common-macro = { workspace = true }
 common-recordbatch = { workspace = true }
 common-runtime = { workspace = true }
 common-telemetry = { workspace = true }

src/common/grpc/src/channel_manager.rs

@@ -16,6 +16,7 @@ use std::sync::atomic::{AtomicBool, AtomicU64, AtomicUsize, Ordering};
 use std::sync::Arc;
 use std::time::Duration;
 
+use common_base::readable_size::ReadableSize;
 use common_telemetry::info;
 use dashmap::mapref::entry::Entry;
 use dashmap::DashMap;
@@ -30,9 +31,9 @@ use crate::error::{CreateChannelSnafu, InvalidConfigFilePathSnafu, InvalidTlsCon
 
 const RECYCLE_CHANNEL_INTERVAL_SECS: u64 = 60;
 pub const DEFAULT_GRPC_REQUEST_TIMEOUT_SECS: u64 = 10;
-pub const DEFAULT_GRPC_CONNECT_TIMEOUT_SECS: u64 = 10;
-pub const DEFAULT_MAX_GRPC_RECV_MESSAGE_SIZE: usize = 512 * 1024 * 1024;
-pub const DEFAULT_MAX_GRPC_SEND_MESSAGE_SIZE: usize = 512 * 1024 * 1024;
+pub const DEFAULT_GRPC_CONNECT_TIMEOUT_SECS: u64 = 1;
+pub const DEFAULT_MAX_GRPC_RECV_MESSAGE_SIZE: ReadableSize = ReadableSize::mb(512);
+pub const DEFAULT_MAX_GRPC_SEND_MESSAGE_SIZE: ReadableSize = ReadableSize::mb(512);
 
 lazy_static! {
     static ref ID: AtomicU64 = AtomicU64::new(0);
@@ -250,9 +251,9 @@ pub struct ChannelConfig {
     pub tcp_nodelay: bool,
     pub client_tls: Option<ClientTlsOption>,
     // Max gRPC receiving(decoding) message size
-    pub max_recv_message_size: usize,
+    pub max_recv_message_size: ReadableSize,
     // Max gRPC sending(encoding) message size
-    pub max_send_message_size: usize,
+    pub max_send_message_size: ReadableSize,
 }
 
 impl Default for ChannelConfig {

src/common/grpc/src/error.rs

@@ -17,19 +17,22 @@ use std::io;
 
 use common_error::ext::ErrorExt;
 use common_error::status_code::StatusCode;
+use common_macro::stack_trace_debug;
 use snafu::{Location, Snafu};
 
 pub type Result<T> = std::result::Result<T, Error>;
 
-#[derive(Debug, Snafu)]
+#[derive(Snafu)]
 #[snafu(visibility(pub))]
+#[stack_trace_debug]
 pub enum Error {
     #[snafu(display("Invalid client tls config, {}", msg))]
     InvalidTlsConfig { msg: String },
 
     #[snafu(display("Invalid config file path"))]
     InvalidConfigFilePath {
-        source: io::Error,
+        #[snafu(source)]
+        error: io::Error,
         location: Location,
     },
 
@@ -48,7 +51,8 @@ pub enum Error {
 
     #[snafu(display("Failed to create gRPC channel"))]
     CreateChannel {
-        source: tonic::transport::Error,
+        #[snafu(source)]
+        error: tonic::transport::Error,
         location: Location,
     },
 
@@ -63,7 +67,8 @@ pub enum Error {
 
     #[snafu(display("Failed to decode FlightData"))]
     DecodeFlightData {
-        source: api::DecodeError,
+        #[snafu(source)]
+        error: api::DecodeError,
         location: Location,
     },
 

src/common/macro/Cargo.toml

@@ -13,6 +13,15 @@ common-telemetry = { workspace = true }
 proc-macro2 = "1.0.66"
 quote = "1.0"
 syn = "1.0"
+syn2 = { version = "2.0", package = "syn", features = [
+    "derive",
+    "parsing",
+    "printing",
+    "clone-impls",
+    "proc-macro",
+    "extra-traits",
+    "full",
+] }
 
 [dev-dependencies]
 arc-swap = "1.0"

src/common/macro/src/lib.rs

@@ -15,6 +15,7 @@
 mod aggr_func;
 mod print_caller;
 mod range_fn;
+mod stack_trace_debug;
 
 use aggr_func::{impl_aggr_func_type_store, impl_as_aggr_func_creator};
 use print_caller::process_print_caller;
@@ -87,3 +88,23 @@ pub fn range_fn(args: TokenStream, input: TokenStream) -> TokenStream {
 pub fn print_caller(args: TokenStream, input: TokenStream) -> TokenStream {
     process_print_caller(args, input)
 }
+
+/// Attribute macro to derive [std::fmt::Debug] for the annotated `Error` type.
+///
+/// The generated `Debug` implementation will print the error in a stack trace style. E.g.:
+/// ```plaintext
+/// 0: Foo error, at src/common/catalog/src/error.rs:80:10
+/// 1: Bar error, at src/common/function/src/error.rs:90:10
+/// 2: Root cause, invalid table name, at src/common/catalog/src/error.rs:100:10
+/// ```
+///
+/// Notes on using this macro:
+/// - `#[snafu(display)]` must present on each enum variants,
+///   and should not include `location` and `source`.
+/// - Only our internal error can be named `source`.
+///   All external error should be `error` with an `#[snafu(source)]` annotation.
+/// - `common_error` crate must be accessible.
+#[proc_macro_attribute]
+pub fn stack_trace_debug(args: TokenStream, input: TokenStream) -> TokenStream {
+    stack_trace_debug::stack_trace_style_impl(args.into(), input.into()).into()
+}

src/common/macro/src/stack_trace_debug.rs

@@ -0,0 +1,278 @@
+// 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.
+
+//! implement `::common_error::ext::StackError`
+
+use proc_macro2::{Span, TokenStream as TokenStream2};
+use quote::{quote, quote_spanned};
+use syn2::spanned::Spanned;
+use syn2::{parenthesized, Attribute, Ident, ItemEnum, Variant};
+
+pub fn stack_trace_style_impl(args: TokenStream2, input: TokenStream2) -> TokenStream2 {
+    let input_cloned: TokenStream2 = input.clone();
+
+    let error_enum_definition: ItemEnum = syn2::parse2(input_cloned).unwrap();
+    let enum_name = error_enum_definition.ident;
+
+    let mut variants = vec![];
+
+    for error_variant in error_enum_definition.variants {
+        let variant = ErrorVariant::from_enum_variant(error_variant);
+        variants.push(variant);
+    }
+
+    let debug_fmt_fn = build_debug_fmt_impl(enum_name.clone(), variants.clone());
+    let next_fn = build_next_impl(enum_name.clone(), variants);
+    let debug_impl = build_debug_impl(enum_name.clone());
+
+    quote! {
+        #args
+        #input
+
+        impl ::common_error::ext::StackError for #enum_name {
+            #debug_fmt_fn
+            #next_fn
+        }
+
+        #debug_impl
+    }
+}
+
+/// Generate `debug_fmt` fn.
+///
+/// The generated fn will be like:
+/// ```rust, ignore
+/// fn debug_fmt(&self, layer: usize, buf: &mut Vec<String>);
+/// ```
+fn build_debug_fmt_impl(enum_name: Ident, variants: Vec<ErrorVariant>) -> TokenStream2 {
+    let match_arms = variants
+        .iter()
+        .map(|v| v.to_debug_match_arm())
+        .collect::<Vec<_>>();
+
+    quote! {
+        fn debug_fmt(&self, layer: usize, buf: &mut Vec<String>) {
+            use #enum_name::*;
+            match self {
+                #(#match_arms)*
+            }
+        }
+    }
+}
+
+/// Generate `next` fn.
+///
+/// The generated fn will be like:
+/// ```rust, ignore
+/// fn next(&self) -> Option<&dyn ::common_error::ext::StackError>;
+/// ```
+fn build_next_impl(enum_name: Ident, variants: Vec<ErrorVariant>) -> TokenStream2 {
+    let match_arms = variants
+        .iter()
+        .map(|v| v.to_next_match_arm())
+        .collect::<Vec<_>>();
+
+    quote! {
+        fn next(&self) -> Option<&dyn ::common_error::ext::StackError> {
+            use #enum_name::*;
+            match self {
+                #(#match_arms)*
+            }
+        }
+    }
+}
+
+/// Implement [std::fmt::Debug] via `debug_fmt`
+fn build_debug_impl(enum_name: Ident) -> TokenStream2 {
+    quote! {
+        impl std::fmt::Debug for #enum_name {
+            fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
+                use ::common_error::ext::StackError;
+                let mut buf = vec![];
+                self.debug_fmt(0, &mut buf);
+                write!(f, "{}", buf.join("\n"))
+            }
+        }
+    }
+}
+
+#[derive(Clone, Debug)]
+struct ErrorVariant {
+    name: Ident,
+    fields: Vec<Ident>,
+    has_location: bool,
+    has_source: bool,
+    has_external_cause: bool,
+    display: TokenStream2,
+    span: Span,
+    cfg_attr: Option<Attribute>,
+}
+
+impl ErrorVariant {
+    /// Construct self from [Variant]
+    fn from_enum_variant(variant: Variant) -> Self {
+        let span = variant.span();
+        let mut has_location = false;
+        let mut has_source = false;
+        let mut has_external_cause = false;
+
+        for field in &variant.fields {
+            if let Some(ident) = &field.ident {
+                if ident == "location" {
+                    has_location = true;
+                } else if ident == "source" {
+                    has_source = true;
+                } else if ident == "error" {
+                    has_external_cause = true;
+                }
+            }
+        }
+
+        let mut display = None;
+        let mut cfg_attr = None;
+        for attr in variant.attrs {
+            if attr.path().is_ident("snafu") {
+                attr.parse_nested_meta(|meta| {
+                    if meta.path.is_ident("display") {
+                        let content;
+                        parenthesized!(content in meta.input);
+                        let display_ts: TokenStream2 = content.parse()?;
+                        display = Some(display_ts);
+                        Ok(())
+                    } else {
+                        Err(meta.error("unrecognized repr"))
+                    }
+                })
+                .expect("Each error should contains a display attribute");
+            }
+
+            if attr.path().is_ident("cfg") {
+                cfg_attr = Some(attr);
+            }
+        }
+
+        let field_ident = variant
+            .fields
+            .iter()
+            .map(|f| f.ident.clone().unwrap_or_else(|| Ident::new("_", f.span())))
+            .collect();
+
+        Self {
+            name: variant.ident,
+            fields: field_ident,
+            has_location,
+            has_source,
+            has_external_cause,
+            display: display.unwrap(),
+            span,
+            cfg_attr,
+        }
+    }
+
+    /// Convert self into an match arm that will be used in [build_debug_impl].
+    ///
+    /// The generated match arm will be like:
+    /// ```rust, ignore
+    ///     ErrorKindWithSource { source, .. } => {
+    ///         debug_fmt(source, layer + 1, buf);
+    ///     },
+    ///     ErrorKindWithoutSource { .. } => {
+    ///        buf.push(format!("{layer}: {}, at {}", format!(#display), location)));
+    ///     }
+    /// ```
+    ///
+    /// The generated code assumes fn `debug_fmt`, var `layer`, var `buf` are in scope.
+    fn to_debug_match_arm(&self) -> TokenStream2 {
+        let name = &self.name;
+        let fields = &self.fields;
+        let display = &self.display;
+        let cfg = if let Some(cfg) = &self.cfg_attr {
+            quote_spanned!(cfg.span() => #cfg)
+        } else {
+            quote! {}
+        };
+
+        match (self.has_location, self.has_source, self.has_external_cause) {
+            (true, true, _) => quote_spanned! {
+               self.span => #cfg #[allow(unused_variables)] #name { #(#fields),*, } => {
+                    buf.push(format!("{layer}: {}, at {}", format!(#display), location));
+                    source.debug_fmt(layer + 1, buf);
+                },
+            },
+            (true, false, true) => quote_spanned! {
+                self.span => #cfg #[allow(unused_variables)] #name { #(#fields),* } => {
+                    buf.push(format!("{layer}: {}, at {}", format!(#display), location));
+                    buf.push(format!("{}: {:?}", layer + 1, error));
+                },
+            },
+            (true, false, false) => quote_spanned! {
+                self.span => #cfg #[allow(unused_variables)] #name { #(#fields),* } => {
+                    buf.push(format!("{layer}: {}, at {}", format!(#display), location));
+                },
+            },
+            (false, true, _) => quote_spanned! {
+                self.span => #cfg #[allow(unused_variables)] #name { #(#fields),* } => {
+                    buf.push(format!("{layer}: {}", format!(#display)));
+                    source.debug_fmt(layer + 1, buf);
+                },
+            },
+            (false, false, true) => quote_spanned! {
+                self.span => #cfg #[allow(unused_variables)] #name { #(#fields),* } => {
+                    buf.push(format!("{layer}: {}", format!(#display)));
+                    buf.push(format!("{}: {:?}", layer + 1, error));
+                },
+            },
+            (false, false, false) => quote_spanned! {
+                self.span => #cfg #[allow(unused_variables)] #name { #(#fields),* } => {
+                    buf.push(format!("{layer}: {}", format!(#display)));
+                },
+            },
+        }
+    }
+
+    /// Convert self into an match arm that will be used in [build_next_impl].
+    ///
+    /// The generated match arm will be like:
+    /// ```rust, ignore
+    ///     ErrorKindWithSource { source, .. } => {
+    ///         Some(source)
+    ///     },
+    ///     ErrorKindWithoutSource { .. } => {
+    ///        None
+    ///     }
+    /// ```
+    fn to_next_match_arm(&self) -> TokenStream2 {
+        let name = &self.name;
+        let fields = &self.fields;
+        let cfg = if let Some(cfg) = &self.cfg_attr {
+            quote_spanned!(cfg.span() => #cfg)
+        } else {
+            quote! {}
+        };
+
+        if self.has_source {
+            quote_spanned! {
+                self.span => #cfg #[allow(unused_variables)] #name { #(#fields),* } => {
+                    Some(source)
+                },
+            }
+        } else {
+            quote_spanned! {
+                self.span => #cfg #[allow(unused_variables)] #name { #(#fields),* } =>{
+                    None
+                }
+            }
+        }
+    }
+}

src/common/mem-prof/Cargo.toml

@@ -6,6 +6,7 @@ license.workspace = true
 
 [dependencies]
 common-error = { workspace = true }
+common-macro = { workspace = true }
 snafu.workspace = true
 tempfile = "3.4"
 tokio.workspace = true

src/common/mem-prof/src/error.rs

@@ -16,14 +16,16 @@ use std::any::Any;
 
 use common_error::ext::{BoxedError, ErrorExt};
 use common_error::status_code::StatusCode;
+use common_macro::stack_trace_debug;
 use snafu::Snafu;
 
 pub type Result<T> = std::result::Result<T, Error>;
 
-#[derive(Debug, Snafu)]
+#[derive(Snafu)]
 #[snafu(visibility(pub))]
+#[stack_trace_debug]
 pub enum Error {
-    #[snafu(display(""))]
+    #[snafu(display("Internal error"))]
     Internal { source: BoxedError },
 
     #[snafu(display("Memory profiling is not supported"))]

src/common/mem-prof/src/jemalloc/error.rs

@@ -17,13 +17,18 @@ use std::path::PathBuf;
 
 use common_error::ext::{BoxedError, ErrorExt};
 use common_error::status_code::StatusCode;
+use common_macro::stack_trace_debug;
 use snafu::{Location, Snafu};
 
-#[derive(Debug, Snafu)]
+#[derive(Snafu)]
 #[snafu(visibility(pub))]
+#[stack_trace_debug]
 pub enum Error {
     #[snafu(display("Failed to read OPT_PROF"))]
-    ReadOptProf { source: tikv_jemalloc_ctl::Error },
+    ReadOptProf {
+        #[snafu(source)]
+        error: tikv_jemalloc_ctl::Error,
+    },
 
     #[snafu(display("Memory profiling is not enabled"))]
     ProfilingNotEnabled,
@@ -34,13 +39,15 @@ pub enum Error {
     #[snafu(display("Failed to open temp file: {}", path))]
     OpenTempFile {
         path: String,
-        source: std::io::Error,
+        #[snafu(source)]
+        error: std::io::Error,
     },
 
     #[snafu(display("Failed to dump profiling data to temp file: {:?}", path))]
     DumpProfileData {
         path: PathBuf,
-        source: tikv_jemalloc_ctl::Error,
+        #[snafu(source)]
+        error: tikv_jemalloc_ctl::Error,
     },
 }
 

src/common/meta/Cargo.toml

@@ -12,9 +12,12 @@ api = { workspace = true }
 arrow-flight.workspace = true
 async-stream.workspace = true
 async-trait.workspace = true
+base64 = "0.21"
+bytes = "1.4"
 common-catalog = { workspace = true }
 common-error = { workspace = true }
 common-grpc-expr.workspace = true
+common-macro = { workspace = true }
 common-procedure = { workspace = true }
 common-recordbatch = { workspace = true }
 common-runtime = { workspace = true }

src/common/meta/src/cache_invalidator.rs

@@ -17,7 +17,6 @@ use std::sync::Arc;
 use table::metadata::TableId;
 
 use crate::error::Result;
-use crate::key::schema_name::SchemaNameKey;
 use crate::key::table_info::TableInfoKey;
 use crate::key::table_name::TableNameKey;
 use crate::key::table_route::TableRouteKey;
@@ -68,36 +67,25 @@ impl CacheInvalidator for DummyCacheInvalidator {
     }
 }
 
-#[derive(Clone)]
-pub struct TableMetadataCacheInvalidator(KvCacheInvalidatorRef);
-
-impl TableMetadataCacheInvalidator {
-    pub fn new(kv_cache_invalidator: KvCacheInvalidatorRef) -> Self {
-        Self(kv_cache_invalidator)
-    }
-
-    pub async fn invalidate_schema(&self, catalog: &str, schema: &str) {
-        let key = SchemaNameKey::new(catalog, schema).as_raw_key();
-        self.0.invalidate_key(&key).await;
-    }
-}
-
 #[async_trait::async_trait]
-impl CacheInvalidator for TableMetadataCacheInvalidator {
+impl<T> CacheInvalidator for T
+where
+    T: KvCacheInvalidator,
+{
     async fn invalidate_table_name(&self, _ctx: &Context, table_name: TableName) -> Result<()> {
         let key: TableNameKey = (&table_name).into();
 
-        self.0.invalidate_key(&key.as_raw_key()).await;
+        self.invalidate_key(&key.as_raw_key()).await;
 
         Ok(())
     }
 
     async fn invalidate_table_id(&self, _ctx: &Context, table_id: TableId) -> Result<()> {
         let key = TableInfoKey::new(table_id);
-        self.0.invalidate_key(&key.as_raw_key()).await;
+        self.invalidate_key(&key.as_raw_key()).await;
 
         let key = &TableRouteKey { table_id };
-        self.0.invalidate_key(&key.as_raw_key()).await;
+        self.invalidate_key(&key.as_raw_key()).await;
 
         Ok(())
     }

src/common/meta/src/ddl.rs

@@ -28,6 +28,7 @@ use crate::rpc::router::RegionRoute;
 pub mod alter_table;
 pub mod create_table;
 pub mod drop_table;
+pub mod truncate_table;
 pub mod utils;
 
 #[derive(Debug, Default)]

src/common/meta/src/ddl/alter_table.rs

@@ -45,6 +45,7 @@ use crate::error::{
 use crate::key::table_info::TableInfoValue;
 use crate::key::table_name::TableNameKey;
 use crate::key::table_route::TableRouteValue;
+use crate::key::DeserializedValueWithBytes;
 use crate::metrics;
 use crate::rpc::ddl::AlterTableTask;
 use crate::rpc::router::{find_leader_regions, find_leaders};
@@ -63,7 +64,7 @@ impl AlterTableProcedure {
     pub fn new(
         cluster_id: u64,
         task: AlterTableTask,
-        table_info_value: TableInfoValue,
+        table_info_value: DeserializedValueWithBytes<TableInfoValue>,
         context: DdlContext,
     ) -> Result<Self> {
         let alter_kind = task
@@ -191,7 +192,8 @@ impl AlterTableProcedure {
             .await?
             .with_context(|| TableRouteNotFoundSnafu {
                 table_name: table_ref.to_string(),
-            })?;
+            })?
+            .into_inner();
 
         let leaders = find_leaders(&region_routes);
         let mut alter_region_tasks = Vec::with_capacity(leaders.len());
@@ -413,7 +415,7 @@ pub struct AlterTableData {
     state: AlterTableState,
     task: AlterTableTask,
     /// Table info value before alteration.
-    table_info_value: TableInfoValue,
+    table_info_value: DeserializedValueWithBytes<TableInfoValue>,
     cluster_id: u64,
     /// Next column id of the table if the task adds columns to the table.
     next_column_id: Option<ColumnId>,
@@ -422,7 +424,7 @@ pub struct AlterTableData {
 impl AlterTableData {
     pub fn new(
         task: AlterTableTask,
-        table_info_value: TableInfoValue,
+        table_info_value: DeserializedValueWithBytes<TableInfoValue>,
         cluster_id: u64,
         next_column_id: Option<ColumnId>,
     ) -> Self {

src/common/meta/src/ddl/create_table.rs

@@ -199,8 +199,8 @@ impl CreateTableProcedure {
             for request in requests {
                 let request = RegionRequest {
                     header: Some(RegionRequestHeader {
-                        trace_id: 0,
-                        span_id: 0,
+                        trace_id: common_telemetry::trace_id().unwrap_or_default(),
+                        ..Default::default()
                     }),
                     body: Some(request),
                 };

src/common/meta/src/ddl/drop_table.rs

@@ -39,6 +39,7 @@ use crate::error::{self, Result};
 use crate::key::table_info::TableInfoValue;
 use crate::key::table_name::TableNameKey;
 use crate::key::table_route::TableRouteValue;
+use crate::key::DeserializedValueWithBytes;
 use crate::metrics;
 use crate::rpc::ddl::DropTableTask;
 use crate::rpc::router::{find_leader_regions, find_leaders, RegionRoute};
@@ -55,8 +56,8 @@ impl DropTableProcedure {
     pub fn new(
         cluster_id: u64,
         task: DropTableTask,
-        table_route_value: TableRouteValue,
-        table_info_value: TableInfoValue,
+        table_route_value: DeserializedValueWithBytes<TableRouteValue>,
+        table_info_value: DeserializedValueWithBytes<TableInfoValue>,
         context: DdlContext,
     ) -> Self {
         Self {
@@ -156,8 +157,8 @@ impl DropTableProcedure {
 
                 let request = RegionRequest {
                     header: Some(RegionRequestHeader {
-                        trace_id: 0,
-                        span_id: 0,
+                        trace_id: common_telemetry::trace_id().unwrap_or_default(),
+                        ..Default::default()
                     }),
                     body: Some(region_request::Body::Drop(PbDropRegionRequest {
                         region_id: region_id.as_u64(),
@@ -231,16 +232,16 @@ pub struct DropTableData {
     pub state: DropTableState,
     pub cluster_id: u64,
     pub task: DropTableTask,
-    pub table_route_value: TableRouteValue,
-    pub table_info_value: TableInfoValue,
+    pub table_route_value: DeserializedValueWithBytes<TableRouteValue>,
+    pub table_info_value: DeserializedValueWithBytes<TableInfoValue>,
 }
 
 impl DropTableData {
     pub fn new(
         cluster_id: u64,
         task: DropTableTask,
-        table_route_value: TableRouteValue,
-        table_info_value: TableInfoValue,
+        table_route_value: DeserializedValueWithBytes<TableRouteValue>,
+        table_info_value: DeserializedValueWithBytes<TableInfoValue>,
     ) -> Self {
         Self {
             state: DropTableState::Prepare,

src/common/meta/src/ddl/truncate_table.rs

@@ -0,0 +1,235 @@
+// 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::region::{
+    region_request, RegionRequest, RegionRequestHeader, TruncateRequest as PbTruncateRegionRequest,
+};
+use async_trait::async_trait;
+use common_procedure::error::{FromJsonSnafu, ToJsonSnafu};
+use common_procedure::{
+    Context as ProcedureContext, LockKey, Procedure, Result as ProcedureResult, Status,
+};
+use common_telemetry::debug;
+use futures::future::join_all;
+use serde::{Deserialize, Serialize};
+use snafu::{ensure, ResultExt};
+use store_api::storage::RegionId;
+use strum::AsRefStr;
+use table::engine::TableReference;
+use table::metadata::{RawTableInfo, TableId};
+
+use super::utils::handle_retry_error;
+use crate::ddl::utils::handle_operate_region_error;
+use crate::ddl::DdlContext;
+use crate::error::{Result, TableNotFoundSnafu};
+use crate::key::table_info::TableInfoValue;
+use crate::key::table_name::TableNameKey;
+use crate::key::DeserializedValueWithBytes;
+use crate::metrics;
+use crate::rpc::ddl::TruncateTableTask;
+use crate::rpc::router::{find_leader_regions, find_leaders, RegionRoute};
+use crate::table_name::TableName;
+
+pub struct TruncateTableProcedure {
+    context: DdlContext,
+    data: TruncateTableData,
+}
+
+#[async_trait]
+impl Procedure for TruncateTableProcedure {
+    fn type_name(&self) -> &str {
+        Self::TYPE_NAME
+    }
+
+    async fn execute(&mut self, _ctx: &ProcedureContext) -> ProcedureResult<Status> {
+        let state = &self.data.state;
+
+        let _timer = common_telemetry::timer!(
+            metrics::METRIC_META_PROCEDURE_TRUNCATE_TABLE,
+            &[("step", state.as_ref().to_string())]
+        );
+
+        match self.data.state {
+            TruncateTableState::Prepare => self.on_prepare().await,
+            TruncateTableState::DatanodeTruncateRegions => {
+                self.on_datanode_truncate_regions().await
+            }
+        }
+        .map_err(handle_retry_error)
+    }
+
+    fn dump(&self) -> ProcedureResult<String> {
+        serde_json::to_string(&self.data).context(ToJsonSnafu)
+    }
+
+    fn lock_key(&self) -> LockKey {
+        let table_ref = &self.data.table_ref();
+        let key = common_catalog::format_full_table_name(
+            table_ref.catalog,
+            table_ref.schema,
+            table_ref.table,
+        );
+
+        LockKey::single(key)
+    }
+}
+
+impl TruncateTableProcedure {
+    pub(crate) const TYPE_NAME: &'static str = "metasrv-procedure::TruncateTable";
+
+    pub(crate) fn new(
+        cluster_id: u64,
+        task: TruncateTableTask,
+        table_info_value: DeserializedValueWithBytes<TableInfoValue>,
+        region_routes: Vec<RegionRoute>,
+        context: DdlContext,
+    ) -> Self {
+        Self {
+            context,
+            data: TruncateTableData::new(cluster_id, task, table_info_value, region_routes),
+        }
+    }
+
+    pub(crate) fn from_json(json: &str, context: DdlContext) -> ProcedureResult<Self> {
+        let data = serde_json::from_str(json).context(FromJsonSnafu)?;
+        Ok(Self { context, data })
+    }
+
+    // Checks whether the table exists.
+    async fn on_prepare(&mut self) -> Result<Status> {
+        let table_ref = &self.data.table_ref();
+
+        let manager = &self.context.table_metadata_manager;
+
+        let exist = manager
+            .table_name_manager()
+            .exists(TableNameKey::new(
+                table_ref.catalog,
+                table_ref.schema,
+                table_ref.table,
+            ))
+            .await?;
+
+        ensure!(
+            exist,
+            TableNotFoundSnafu {
+                table_name: table_ref.to_string()
+            }
+        );
+
+        self.data.state = TruncateTableState::DatanodeTruncateRegions;
+
+        Ok(Status::executing(true))
+    }
+
+    async fn on_datanode_truncate_regions(&mut self) -> Result<Status> {
+        let table_id = self.data.table_id();
+
+        let region_routes = &self.data.region_routes;
+        let leaders = find_leaders(region_routes);
+        let mut truncate_region_tasks = Vec::with_capacity(leaders.len());
+
+        for datanode in leaders {
+            let requester = self.context.datanode_manager.datanode(&datanode).await;
+            let regions = find_leader_regions(region_routes, &datanode);
+
+            for region in regions {
+                let region_id = RegionId::new(table_id, region);
+                debug!(
+                    "Truncating table {} region {} on Datanode {:?}",
+                    self.data.table_ref(),
+                    region_id,
+                    datanode
+                );
+
+                let request = RegionRequest {
+                    header: Some(RegionRequestHeader {
+                        trace_id: common_telemetry::trace_id().unwrap_or_default(),
+                        ..Default::default()
+                    }),
+                    body: Some(region_request::Body::Truncate(PbTruncateRegionRequest {
+                        region_id: region_id.as_u64(),
+                    })),
+                };
+
+                let datanode = datanode.clone();
+                let requester = requester.clone();
+
+                truncate_region_tasks.push(async move {
+                    if let Err(err) = requester.handle(request).await {
+                        return Err(handle_operate_region_error(datanode)(err));
+                    }
+                    Ok(())
+                });
+            }
+        }
+
+        join_all(truncate_region_tasks)
+            .await
+            .into_iter()
+            .collect::<Result<Vec<_>>>()?;
+
+        Ok(Status::Done)
+    }
+}
+
+#[derive(Debug, Serialize, Deserialize)]
+pub struct TruncateTableData {
+    state: TruncateTableState,
+    cluster_id: u64,
+    task: TruncateTableTask,
+    table_info_value: DeserializedValueWithBytes<TableInfoValue>,
+    region_routes: Vec<RegionRoute>,
+}
+
+impl TruncateTableData {
+    pub fn new(
+        cluster_id: u64,
+        task: TruncateTableTask,
+        table_info_value: DeserializedValueWithBytes<TableInfoValue>,
+        region_routes: Vec<RegionRoute>,
+    ) -> Self {
+        Self {
+            state: TruncateTableState::Prepare,
+            cluster_id,
+            task,
+            table_info_value,
+            region_routes,
+        }
+    }
+
+    pub fn table_ref(&self) -> TableReference {
+        self.task.table_ref()
+    }
+
+    pub fn table_name(&self) -> TableName {
+        self.task.table_name()
+    }
+
+    fn table_info(&self) -> &RawTableInfo {
+        &self.table_info_value.table_info
+    }
+
+    fn table_id(&self) -> TableId {
+        self.table_info().ident.table_id
+    }
+}
+
+#[derive(Debug, Serialize, Deserialize, AsRefStr)]
+enum TruncateTableState {
+    /// Prepares to truncate the table
+    Prepare,
+    /// Truncates regions on Datanode
+    DatanodeTruncateRegions,
+}

src/common/meta/src/ddl_manager.rs

@@ -15,7 +15,7 @@
 use std::sync::Arc;
 
 use common_procedure::{watcher, ProcedureId, ProcedureManagerRef, ProcedureWithId};
-use common_telemetry::{error, info};
+use common_telemetry::info;
 use snafu::{OptionExt, ResultExt};
 
 use crate::cache_invalidator::CacheInvalidatorRef;
@@ -23,18 +23,19 @@ use crate::datanode_manager::DatanodeManagerRef;
 use crate::ddl::alter_table::AlterTableProcedure;
 use crate::ddl::create_table::CreateTableProcedure;
 use crate::ddl::drop_table::DropTableProcedure;
+use crate::ddl::truncate_table::TruncateTableProcedure;
 use crate::ddl::{
     DdlContext, DdlTaskExecutor, ExecutorContext, TableMetadataAllocatorContext,
     TableMetadataAllocatorRef,
 };
 use crate::error::{
     self, RegisterProcedureLoaderSnafu, Result, SubmitProcedureSnafu, TableNotFoundSnafu,
-    UnsupportedSnafu, WaitProcedureSnafu,
+    WaitProcedureSnafu,
 };
 use crate::key::table_info::TableInfoValue;
 use crate::key::table_name::TableNameKey;
 use crate::key::table_route::TableRouteValue;
-use crate::key::TableMetadataManagerRef;
+use crate::key::{DeserializedValueWithBytes, TableMetadataManagerRef};
 use crate::rpc::ddl::DdlTask::{AlterTable, CreateTable, DropTable, TruncateTable};
 use crate::rpc::ddl::{
     AlterTableTask, CreateTableTask, DropTableTask, SubmitDdlTaskRequest, SubmitDdlTaskResponse,
@@ -122,6 +123,20 @@ impl DdlManager {
             )
             .context(RegisterProcedureLoaderSnafu {
                 type_name: AlterTableProcedure::TYPE_NAME,
+            })?;
+
+        let context = self.create_context();
+
+        self.procedure_manager
+            .register_loader(
+                TruncateTableProcedure::TYPE_NAME,
+                Box::new(move |json| {
+                    let context = context.clone();
+                    TruncateTableProcedure::from_json(json, context).map(|p| Box::new(p) as _)
+                }),
+            )
+            .context(RegisterProcedureLoaderSnafu {
+                type_name: TruncateTableProcedure::TYPE_NAME,
             })
     }
 
@@ -129,7 +144,7 @@ impl DdlManager {
         &self,
         cluster_id: u64,
         alter_table_task: AlterTableTask,
-        table_info_value: TableInfoValue,
+        table_info_value: DeserializedValueWithBytes<TableInfoValue>,
     ) -> Result<ProcedureId> {
         let context = self.create_context();
 
@@ -161,8 +176,8 @@ impl DdlManager {
         &self,
         cluster_id: u64,
         drop_table_task: DropTableTask,
-        table_info_value: TableInfoValue,
-        table_route_value: TableRouteValue,
+        table_info_value: DeserializedValueWithBytes<TableInfoValue>,
+        table_route_value: DeserializedValueWithBytes<TableRouteValue>,
     ) -> Result<ProcedureId> {
         let context = self.create_context();
 
@@ -183,15 +198,21 @@ impl DdlManager {
         &self,
         cluster_id: u64,
         truncate_table_task: TruncateTableTask,
+        table_info_value: DeserializedValueWithBytes<TableInfoValue>,
         region_routes: Vec<RegionRoute>,
     ) -> Result<ProcedureId> {
-        error!("Truncate table procedure is not supported, cluster_id = {}, truncate_table_task = {:?}, region_routes = {:?}",
-            cluster_id, truncate_table_task, region_routes);
+        let context = self.create_context();
+        let procedure = TruncateTableProcedure::new(
+            cluster_id,
+            truncate_table_task,
+            table_info_value,
+            region_routes,
+            context,
+        );
 
-        UnsupportedSnafu {
-            operation: "TRUNCATE TABLE",
-        }
-        .fail()
+        let procedure_with_id = ProcedureWithId::with_random_id(Box::new(procedure));
+
+        self.submit_procedure(procedure_with_id).await
     }
 
     async fn submit_procedure(&self, procedure_with_id: ProcedureWithId) -> Result<ProcedureId> {
@@ -216,32 +237,34 @@ async fn handle_truncate_table_task(
     cluster_id: u64,
     truncate_table_task: TruncateTableTask,
 ) -> Result<SubmitDdlTaskResponse> {
-    let truncate_table = &truncate_table_task.truncate_table;
-    let table_id = truncate_table
-        .table_id
-        .as_ref()
-        .context(error::UnexpectedSnafu {
-            err_msg: "expected table id ",
-        })?
-        .id;
-
+    let table_id = truncate_table_task.table_id;
+    let table_metadata_manager = &ddl_manager.table_metadata_manager();
     let table_ref = truncate_table_task.table_ref();
 
-    let table_route_value = ddl_manager
-        .table_metadata_manager()
-        .table_route_manager()
-        .get(table_id)
-        .await?
-        .with_context(|| error::TableRouteNotFoundSnafu {
-            table_name: table_ref.to_string(),
-        })?;
+    let (table_info_value, table_route_value) =
+        table_metadata_manager.get_full_table_info(table_id).await?;
 
-    let table_route = table_route_value.region_routes;
+    let table_info_value = table_info_value.with_context(|| error::TableInfoNotFoundSnafu {
+        table_name: table_ref.to_string(),
+    })?;
+
+    let table_route_value = table_route_value.with_context(|| error::TableRouteNotFoundSnafu {
+        table_name: table_ref.to_string(),
+    })?;
+
+    let table_route = table_route_value.into_inner().region_routes;
 
     let id = ddl_manager
-        .submit_truncate_table_task(cluster_id, truncate_table_task, table_route)
+        .submit_truncate_table_task(
+            cluster_id,
+            truncate_table_task,
+            table_info_value,
+            table_route,
+        )
         .await?;
 
+    info!("Table: {table_id} is truncated via procedure_id {id:?}");
+
     Ok(SubmitDdlTaskResponse {
         key: id.to_string().into(),
         ..Default::default()

src/common/meta/src/distributed_time_constants.rs

@@ -29,3 +29,9 @@ pub const REGION_LEASE_SECS: u64 =
 /// When creating table or region failover, a target node needs to be selected.
 /// If the node's lease has expired, the `Selector` will not select it.
 pub const DATANODE_LEASE_SECS: u64 = REGION_LEASE_SECS;
+
+/// The lease seconds of metasrv leader.
+pub const META_LEASE_SECS: u64 = 3;
+
+// In a lease, there are two opportunities for renewal.
+pub const META_KEEP_ALIVE_INTERVAL_SECS: u64 = META_LEASE_SECS / 2;