Add spark 3.4 unit test scripts

apache · Oct 28, 2023 · a0313b6 · a0313b6
1 parent dec4323
commit a0313b6
Showing 1 changed file with 67 additions and 0 deletions.
diff --git a/.github/workflows/velox_be.yml b/.github/workflows/velox_be.yml
@@ -193,6 +193,73 @@ jobs:
         if: ${{ always() }}
         run: |
           docker stop ubuntu2004-test-spark33-$GITHUB_RUN_ID || true
+  
+  ubuntu2004-test-spark34-slow:
+    runs-on: velox-self-hosted
+    steps:
+      - uses: actions/checkout@v2
+      - name: Setup docker container
+        run: |
+          docker run --rm --init --privileged --ulimit nofile=65536:65536 --ulimit core=-1 --security-opt seccomp=unconfined \
+          -v $PWD:/opt/gluten --name ubuntu2004-test-spark34-slow-$GITHUB_RUN_ID -e NUM_THREADS=30 -detach 10.0.2.4:5000/gluten-dev/ubuntu:20.04 \
+          'cd /opt/gluten && sleep 14400'
+      - name: Build Gluten velox third party
+        run: |
+          docker exec ubuntu2004-test-spark34-slow-$GITHUB_RUN_ID bash -l -c '
+          cd /opt/gluten/ep/build-velox/src && \
+          ./get_velox.sh --velox_home=/opt/velox && \
+          ./build_velox.sh --velox_home=/opt/velox --enable_ep_cache=ON'
+      - name: Build Gluten CPP library
+        run: |
+          docker exec ubuntu2004-test-spark34-slow-$GITHUB_RUN_ID bash -l -c '
+          cd /opt/gluten/cpp && \
+          ./compile.sh --build_velox_backend=ON --velox_home=/opt/velox --arrow_home=/opt/velox/_build/release/third_party/arrow_ep'
+      - name: Build and Run unit test for Spark 3.4.1(slow tests)
+        run: |
+          docker exec ubuntu2004-test-spark34-slow-$GITHUB_RUN_ID bash -l -c 'cd /opt/gluten && \
+          mvn clean install -Pspark-3.4 -Pbackends-velox -Prss -Pspark-ut -DargLine="-Dspark.test.home=/opt/spark341" -DtagsToInclude=org.apache.spark.tags.ExtendedSQLTest'
+      - name: TPC-H SF1.0 && TPC-DS SF1.0 Parquet local spark3.4
+        run: |
+          docker exec ubuntu2004-test-spark34-slow-$GITHUB_RUN_ID bash -l -c 'cd /opt/gluten/tools/gluten-it && \
+          mvn clean install -Pspark-3.4 \
+          && GLUTEN_IT_JVM_ARGS=-Xmx5G sbin/gluten-it.sh queries-compare \
+            --local --preset=velox --benchmark-type=h --error-on-memleak --disable-aqe --off-heap-size=10g -s=1.0 --threads=16 --iterations=1 \
+          && GLUTEN_IT_JVM_ARGS=-Xmx5G sbin/gluten-it.sh queries-compare \
+            --local --preset=velox --benchmark-type=ds --error-on-memleak --off-heap-size=10g -s=1.0 --threads=16 --iterations=1'
+      - name: Exit docker container
+        if: ${{ always() }}
+        run: |
+          docker stop ubuntu2004-test-spark34-slow-$GITHUB_RUN_ID || true
+
+  ubuntu2004-test-spark34:
+    runs-on: velox-self-hosted
+    steps:
+      - uses: actions/checkout@v2
+      - name: Setup docker container
+        run: |
+          docker run --rm --init --privileged --ulimit nofile=65536:65536 --ulimit core=-1 --security-opt seccomp=unconfined \
+          -v $PWD:/opt/gluten --name ubuntu2004-test-spark34-$GITHUB_RUN_ID -e NUM_THREADS=30 -detach 10.0.2.4:5000/gluten-dev/ubuntu:20.04 \
+          'cd /opt/gluten && sleep 14400'
+      - name: Build Gluten velox third party
+        run: |
+          docker exec ubuntu2004-test-spark34-$GITHUB_RUN_ID bash -c '
+          cd /opt/gluten/ep/build-velox/src && \
+          ./get_velox.sh --velox_home=/opt/velox && \
+          ./build_velox.sh --velox_home=/opt/velox --enable_ep_cache=ON'
+      - name: Build Gluten CPP library
+        run: |
+          docker exec ubuntu2004-test-spark34-$GITHUB_RUN_ID bash -c '
+          cd /opt/gluten/cpp && \
+          ./compile.sh --build_velox_backend=ON --velox_home=/opt/velox --arrow_home=/opt/velox/_build/release/third_party/arrow_ep --build_examples=ON'
+      - name: Build and Run unit test for Spark 3.4.1(other tests)
+        run: |
+          docker exec ubuntu2004-test-spark34-$GITHUB_RUN_ID bash -c 'cd /opt/gluten && \
+          mvn clean install -Pspark-3.4 -Pbackends-velox -Prss -Pspark-ut -DargLine="-Dspark.test.home=/opt/spark341" -DtagsToExclude=org.apache.spark.tags.ExtendedSQLTest,io.glutenproject.tags.UDFTest,io.glutenproject.tags.SkipTestTags && \
+          mvn test -Pspark-3.4 -Pbackends-velox -DtagsToExclude=None -DtagsToInclude=io.glutenproject.tags.UDFTest'
+      - name: Exit docker container
+        if: ${{ always() }}
+        run: |
+          docker stop ubuntu2004-test-spark34-$GITHUB_RUN_ID || true
 
   ubuntu2204-test:
     runs-on: velox-self-hosted