Skip to content

Apache DataFusion SQL Query Engine

bench: make wide_schema honor DATA_DIR like the other sql_benchmarks (#22836)

Adrian Garcia Badaracco9b81ff86d1ad
diff --git a/benchmarks/bench.sh b/benchmarks/bench.sh
index 16b96da477..0e28beadf8 100755
--- a/benchmarks/bench.sh
+++ b/benchmarks/bench.sh
@@ -786,12 +786,14 @@ data_wide_schema() {
run_wide_schema() {
echo "Running wide_schema benchmark (wide subgroup)..."
debug_run env BENCH_NAME=wide_schema BENCH_SUBGROUP=wide \
+ DATA_DIR="${DATA_DIR}" \
SIMULATE_LATENCY="${SIMULATE_LATENCY}" \
${QUERY:+BENCH_QUERY="${QUERY}"} \
bash -c "$SQL_CARGO_COMMAND"
echo "Running wide_schema benchmark (narrow baseline subgroup)..."
debug_run env BENCH_NAME=wide_schema BENCH_SUBGROUP=narrow \
+ DATA_DIR="${DATA_DIR}" \
SIMULATE_LATENCY="${SIMULATE_LATENCY}" \
${QUERY:+BENCH_QUERY="${QUERY}"} \
bash -c "$SQL_CARGO_COMMAND"
diff --git a/benchmarks/sql_benchmarks/wide_schema/init/load.sql b/benchmarks/sql_benchmarks/wide_schema/init/load.sql
index 4fbcda1d58..7248610677 100644
--- a/benchmarks/sql_benchmarks/wide_schema/init/load.sql
+++ b/benchmarks/sql_benchmarks/wide_schema/init/load.sql
@@ -3,4 +3,4 @@
-- BENCH_SUBGROUP=wide → 1024-col synthetic dataset (the actual benchmark)
-- BENCH_SUBGROUP=narrow → 8-col baseline (companion only — meaningful
-- only when compared to the wide numbers)
-CREATE EXTERNAL TABLE events STORED AS PARQUET LOCATION 'data/wide_schema/${BENCH_SUBGROUP:-wide}/';
+CREATE EXTERNAL TABLE events STORED AS PARQUET LOCATION '${DATA_DIR:-data}/wide_schema/${BENCH_SUBGROUP:-wide}/';