Skip to content

Apache DataFusion SQL Query Engine

Update docs for aggregate repartition test (#18650)

Xander5053c71af6f0
diff --git a/datafusion/sqllogictest/test_files/aggregate_repartition.slt b/datafusion/sqllogictest/test_files/aggregate_repartition.slt
index 7612fc84d4..5db26f00a7 100644
--- a/datafusion/sqllogictest/test_files/aggregate_repartition.slt
+++ b/datafusion/sqllogictest/test_files/aggregate_repartition.slt
@@ -84,9 +84,6 @@ physical_plan
07)------------DataSourceExec: file_groups={1 group: [[WORKSPACE_ROOT/datafusion/sqllogictest/test_files/scratch/aggregate_repartition/dim.csv]]}, projection=[env], file_type=csv, has_header=true
# Test 2: EXPLAIN query for Parquet table with GROUP BY
-# This plan differs from the one above and includes two consecutive repartitions — one round-robin and one hash —
-# which seems unnecessary. We may want to align it with the previous plan (push the round robin down or remove the round robin), or, if the input file is small,
-# avoid repartitioning altogether. A single partition should suffice for a single-step aggregate as the plan after this.
query TT
EXPLAIN SELECT env, count(*) FROM dim_parquet GROUP BY env;