Generate expected benchmark query results (apache#4010)

yuuch · Nov 5, 2022 · c783b9e · c783b9e
1 parent 49058d5
commit c783b9e
Show file tree

Hide file tree

Showing 3 changed files with 12 additions and 59 deletions.
diff --git a/benchmarks/entrypoint.sh b/benchmarks/entrypoint.sh
diff --git a/benchmarks/tpch-gen.sh b/benchmarks/tpch-gen.sh
@@ -16,20 +16,27 @@
 # specific language governing permissions and limitations
 # under the License.
 
-#set -e
+mkdir -p data/answers 2>/dev/null
+
+set -e
 
 pushd ..
 . ./dev/build-set-env.sh
 popd
 
-docker build  -f tpchgen.dockerfile -t datafusion-tpchgen:$DATAFUSION_VERSION .
-
 # Generate data into the ./data directory if it does not already exist
 FILE=./data/supplier.tbl
 if test -f "$FILE"; then
     echo "$FILE exists."
 else
-  mkdir data 2>/dev/null
-  docker run -v `pwd`/data:/data -it --rm datafusion-tpchgen:$DATAFUSION_VERSION $1
+  docker run -v `pwd`/data:/data -it --rm ghcr.io/databloom-ai/tpch-docker:main -vf -s $1
   ls -l data
+fi
+
+# Copy expected answers (at SF=1) into the ./data/answers directory if it does not already exist
+FILE=./data/answers/q1.out
+if test -f "$FILE"; then
+    echo "$FILE exists."
+else
+  docker run -v `pwd`/data:/data -it --entrypoint /bin/bash --rm ghcr.io/databloom-ai/tpch-docker:main -c "cp /opt/tpch/2.18.0_rc2/dbgen/answers/* /data/answers/"
 fi
diff --git a/benchmarks/tpchgen.dockerfile b/benchmarks/tpchgen.dockerfile