Merge branch 'master' into bq-retries

datahub-project · Aug 26, 2024 · 239e163 · 239e163
2 parents 9af1c97 + ca4c436
commit 239e163
Show file tree

Hide file tree

Showing 298 changed files with 40,091 additions and 5,195 deletions.
diff --git a/.github/actions/ci-optimization/action.yml b/.github/actions/ci-optimization/action.yml
@@ -1,5 +1,5 @@
-name: 'Identify CI Optimizations'
-description: 'Determine if code changes are specific to certain modules.'
+name: "Identify CI Optimizations"
+description: "Determine if code changes are specific to certain modules."
 
 outputs:
   frontend-only:
@@ -44,27 +44,25 @@ outputs:
 runs:
   using: "composite"
   steps:
-    - uses: dorny/paths-filter@v2
+    - uses: dorny/paths-filter@v3
       id: filter
       with:
+        token: "" # Empty token forces it to use raw git commands.
         filters: |
           frontend:
             - "datahub-frontend/**"
             - "datahub-web-react/**"
-            - "smoke-test/tests/cypress/**"
             - "docker/datahub-frontend/**"
           ingestion:
             - "metadata-ingestion-modules/**"
             - "metadata-ingestion/**"
             - "metadata-models/**"
-            - "smoke-test/**"
             - "docker/datahub-ingestion**"
           ingestion-base:
             - "docker/datahub-ingestion-base/**"
           docker:
             - "docker/**"
           backend:
-            - ".github/**"
             - "metadata-models/**"
             - "datahub-upgrade/**"
             - "entity-registry/**"
@@ -78,7 +76,6 @@ runs:
             - "metadata-utils/**"
             - "metadata-operation-context/**"
             - "datahub-graphql-core/**"
-            - "smoke-test/**"
             - "docker/**"
           kafka-setup:
             - "docker/kafka-setup/**"

diff --git a/.github/actions/docker-custom-build-and-push/action.yml b/.github/actions/docker-custom-build-and-push/action.yml
@@ -26,10 +26,13 @@ inputs:
   build-args:
     description: "List of build-time variables. Same as docker/build-push-action"
     required: false
-  tags:
-    # e.g. latest,head,sha12345
-    description: "List of tags to use for the Docker image"
+  image_tag:
+    # e.g. pr12345 OR head OR v0.1.2.3
+    description: "Main tag to use for the Docker image"
     required: true
+  flavor:
+    description: 'Image flavor (e.g., slim, full)'
+    required: false
   target:
     description: "Sets the target stage to build"
     required: false
@@ -45,13 +48,16 @@ runs:
   steps:
     - name: Docker meta
       id: docker_meta
-      uses: crazy-max/ghaction-docker-meta@v1
+      uses: docker/metadata-action@v5
       with:
-        # list of Docker images to use as base name for tags
         images: ${{ inputs.images }}
-        # add git short SHA as Docker tag
-        tag-custom: ${{ inputs.tags }}
-        tag-custom-only: true
+        flavor: |
+          latest=false
+          suffix=${{ inputs.flavor && format('-{0}', inputs.flavor) || '' }}
+        tags: |
+          type=raw,value=${{ inputs.image_tag }}
+          type=raw,value=head,enable={{is_default_branch}}
+          type=sha,prefix=,format=short
 
     # Code for testing the build when not pushing to Docker Hub.
     - name: Build and Load image for testing (if not publishing)
@@ -74,10 +80,13 @@ runs:
       if: ${{ inputs.publish != 'true' }}
       shell: bash
       run: |
+        IMAGES="""
+        ${{ inputs.images }}
+        """
         TAGS="""
-        ${{ steps.docker_meta.outputs.tags }}
+        ${{ inputs.image_tag }}
         """
-        echo "SINGLE_TAG=$(echo $TAGS | tr '\n' ' ' | awk -F' ' '{ print $1 }')" >> $GITHUB_OUTPUT
+        echo "SINGLE_TAG=$(echo $IMAGES | tr '\n' ' ' | awk -F' |,' '{ print $1 }'):$(echo $TAGS | tr '\n' ' ' | awk -F' |,' '{ print $1 }')" >> $GITHUB_OUTPUT
       id: single_tag
     - name: Upload image locally for testing (if not publishing)
       uses: ishworkh/docker-image-artifact-upload@v1

diff --git a/.github/scripts/docker_helpers.sh b/.github/scripts/docker_helpers.sh
@@ -5,22 +5,22 @@ export MAIN_BRANCH="master"
 export MAIN_BRANCH_TAG="head"
 
 function get_short_sha {
-    echo $(git rev-parse --short "$GITHUB_SHA")
+    echo $(git rev-parse --short "$GITHUB_SHA"|head -c7)
 }
 
 export SHORT_SHA=$(get_short_sha)
 echo "SHORT_SHA: $SHORT_SHA"
 
 function get_tag {
-    echo $(echo ${GITHUB_REF} | sed -e "s,refs/heads/${MAIN_BRANCH},${MAIN_BRANCH_TAG},g" -e 's,refs/tags/,,g' -e 's,refs/pull/\([0-9]*\).*,pr\1,g'),${SHORT_SHA}
+    echo $(echo ${GITHUB_REF} | sed -e "s,refs/heads/${MAIN_BRANCH},${MAIN_BRANCH_TAG},g" -e 's,refs/tags/,,g' -e 's,refs/pull/\([0-9]*\).*,pr\1,g')
 }
 
 function get_tag_slim {
-    echo $(echo ${GITHUB_REF} | sed -e "s,refs/heads/${MAIN_BRANCH},${MAIN_BRANCH_TAG}-slim,g" -e 's,refs/tags/\(.*\),\1-slim,g' -e 's,refs/pull/\([0-9]*\).*,pr\1-slim,g'),${SHORT_SHA}-slim
+    echo $(echo ${GITHUB_REF} | sed -e "s,refs/heads/${MAIN_BRANCH},${MAIN_BRANCH_TAG}-slim,g" -e 's,refs/tags/\(.*\),\1-slim,g' -e 's,refs/pull/\([0-9]*\).*,pr\1-slim,g')
 }
 
 function get_tag_full {
-    echo $(echo ${GITHUB_REF} | sed -e "s,refs/heads/${MAIN_BRANCH},${MAIN_BRANCH_TAG}-full,g" -e 's,refs/tags/\(.*\),\1-full,g' -e 's,refs/pull/\([0-9]*\).*,pr\1-full,g'),${SHORT_SHA}-full
+    echo $(echo ${GITHUB_REF} | sed -e "s,refs/heads/${MAIN_BRANCH},${MAIN_BRANCH_TAG}-full,g" -e 's,refs/tags/\(.*\),\1-full,g' -e 's,refs/pull/\([0-9]*\).*,pr\1-full,g')
 }
 
 function get_python_docker_release_v {
@@ -38,3 +38,11 @@ function get_unique_tag_slim {
 function get_unique_tag_full {
     echo $(echo ${GITHUB_REF} | sed -e "s,refs/heads/${MAIN_BRANCH},${SHORT_SHA}-full,g" -e 's,refs/tags/\(.*\),\1-full,g' -e 's,refs/pull/\([0-9]*\).*,pr\1-full,g')
 }
+
+function get_platforms_based_on_branch {
+    if [ "${{ github.event_name }}" == 'push' && "${{ github.ref }}" == "refs/heads/${MAIN_BRANCH}" ]; then
+        echo "linux/amd64,linux/arm64"
+    else
+        echo "linux/amd64"
+    fi
+}
diff --git a/.github/scripts/docker_logs.sh b/.github/scripts/docker_logs.sh
@@ -0,0 +1,8 @@
+TARGET_DIR="${TARGET_DIR:=docker_logs}"
+TEST_STRATEGY="${TEST_STRATEGY:=}"
+
+mkdir -p "$TARGET_DIR"
+for name in `docker ps -a --format '{{.Names}}'`;
+do
+    docker logs "$name" >& "${TARGET_DIR}/${name}${TEST_STRATEGY}.log" || true
+done
diff --git a/.github/workflows/build-and-test.yml b/.github/workflows/build-and-test.yml
@@ -57,6 +57,11 @@ jobs:
     timeout-minutes: 60
     needs: setup
     steps:
+      - name: Free up disk space
+        run: |
+          sudo apt-get remove 'dotnet-*' azure-cli || true
+          sudo rm -rf /usr/local/lib/android/ || true
+          sudo docker image prune -a -f || true
       - uses: szenius/[email protected]
         with:
           timezoneLinux: ${{ matrix.timezone }}
@@ -86,6 +91,8 @@ jobs:
             -x :metadata-ingestion-modules:airflow-plugin:check \
             -x :metadata-ingestion-modules:dagster-plugin:build \
             -x :metadata-ingestion-modules:dagster-plugin:check \
+            -x :metadata-ingestion-modules:gx-plugin:build \
+            -x :metadata-ingestion-modules:gx-plugin:check \
             -x :datahub-frontend:build \
             -x :datahub-web-react:build \
             --parallel