ClickBench icon indicating copy to clipboard operation
ClickBench copied to clipboard

WIP (NOT READY FOR REVIEW) Update Results for DataFusion 51.0.0

Open alamb opened this issue 3 months ago • 3 comments

  • Closes https://github.com/apache/datafusion/issues/18909

Update DataFusion results for the DataFusion 51 release (TODO add blog URL here when published)

I followed the directions in

  • https://github.com/ClickHouse/ClickBench/blob/main/datafusion/README.md

Changes

  • [x] Update some readme contents to remove outdated contents
  • [ ] Fix lukewarm-code tagging (the scripts in this repository run datafusion-cli from scratch each time, so there are no caches maintained from query to query) - https://github.com/ClickHouse/ClickBench/pull/692#discussion_r2556531230
  • [ ] Add scripts to convert from csv --> json result format

Variants:

  • [ ] DataFusion parquet
  • [ ] DataFusion parquet-partitioned

Note I did not include datafusion with vortex (TBD ping SpiralDB)

Results included

  • [ ] c6a.4xlarge
  • [ ] c6a.2xlarge
  • [ ] c6a.xlarge
  • [ ] c6a.large

Not sure

  • [ ] c8g.4xlarge
  • [ ] t3a.small

alamb avatar Nov 24 '25 14:11 alamb

CLA assistant check
All committers have signed the CLA.

CLAassistant avatar Nov 24 '25 14:11 CLAassistant

@alamb Please ping me when this PR is ready for review - thanks.

rschu1ze avatar Nov 24 '25 19:11 rschu1ze

Thank you @rschu1ze -- I am still doing some performance analysis (you can see details here if you care https://github.com/apache/datafusion/issues/18909). I will let you know when it is ready

alamb avatar Nov 24 '25 22:11 alamb