Skip to content

Support spilling for Presto on Spark#16133

Merged
arhimondr merged 6 commits intoprestodb:masterfrom
viczhang861:pos_spill_push
May 20, 2021
Merged

Support spilling for Presto on Spark#16133
arhimondr merged 6 commits intoprestodb:masterfrom
viczhang861:pos_spill_push

Conversation

@viczhang861
Copy link
Contributor

This PR only makes spilling work, it is not targeted for memory improvement after enabling spilling.

Test

  • Unit test
    Enable spilling and manually verify spilled file is created.
  • Integration test
    Run large Presto on Spark query and verify data is spilled in logging.
== RELEASE NOTES ==

Presto on Spark
* Add session property `spark_memory_revoking_threshold` and configuration property `spark.memory-revoking-threshold`, spilling is triggered when total memory is beyond this threshold.

@viczhang861 viczhang861 requested a review from a team May 20, 2021 15:12
@viczhang861
Copy link
Contributor Author

There is one flaky test testEmptyJoins, disabled and created #16135 to track

Revocalbe memory should be counted towards total used
memory for Presto on Spark query.
Add revocable memory to node total memory for
Presto on Spark query only.
Tuning memory revoking threshold by session property
is only allowed for Presto no Spark query.
@arhimondr arhimondr merged commit e2944c6 into prestodb:master May 20, 2021
@viczhang861 viczhang861 deleted the pos_spill_push branch May 21, 2021 01:07
@sujay-jain sujay-jain mentioned this pull request May 21, 2021
10 tasks
@viczhang861 viczhang861 restored the pos_spill_push branch June 15, 2021 21:08
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants