ErikBPF opened a new pull request, #5904:
URL: https://github.com/apache/datafusion-comet/pull/5904
## Which issue does this PR close?
Related to #5704. This restores the previous suite setting; it does not
establish whether other workloads have changed memory requirements.
## Rationale for this change
The SF1 TPC-H suite passes on current main with the original 2 GiB off-heap
budget, so the increase to 4 GiB is no longer needed for this configuration.
## What changes are included in this PR?
Restore `MEMORY_OFFHEAP_SIZE` from `4g` to `2g` in `CometTPCHQuerySuite`.
Query plans, data generation, and expected results are unchanged.
## How are these changes tested?
On base `f69c4c81b9429e327ea95658530ae4ed4ed19635`, Spark 4.1.3 / Scala
2.13.17 / JDK 17, the complete suite passed against repository-generated
SF1 Parquet with one partition per table, with both debug and release
native libraries: 22 tests, 66 query/join combinations, zero ignored.
```bash
SPARK_TPCH_DATA=/path/to/sf1_parquet ./mvnw -B -Prelease test \
-Dtest=none -Dsuites=org.apache.spark.sql.CometTPCHQuerySuite
```
This is a suite-budget check, not a historical performance or memory
benchmark. Other Spark versions and the original DataFusion upgrade commit
were not measured.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
