Skip to content

SNOW-2380741: Use all logical cores in snowpark tests.#3824

Merged
sfc-gh-mvashishtha merged 8 commits into
mainfrom
mavshishtha/SNOW-2380741/use-all-cores-in-ci
Oct 3, 2025
Merged

SNOW-2380741: Use all logical cores in snowpark tests.#3824
sfc-gh-mvashishtha merged 8 commits into
mainfrom
mavshishtha/SNOW-2380741/use-all-cores-in-ci

Conversation

@sfc-gh-mvashishtha

@sfc-gh-mvashishtha sfc-gh-mvashishtha commented Sep 30, 2025

Copy link
Copy Markdown
Contributor

Formerly, we had limited the test parallelism to not overwhelm our Snowflake deployments, but now that we are using multi-cluster warehouses, we should be able to use all 64 cores.

This change speeds up Modin CI jobs according to a sample size of 1. Other long-running CI jobs seem to take about as long before and after. before vs after:

test before (minutes) after (minutes)
modin-macos-latest-3.9-aws 30 28
modin-ubuntu-latest-64-cores-3.10-azure 42 31
modin-ubuntu-latest-64-cores-3.12-aws 48 30
modin-windows-latest-64-cores-3.11-gcp 71 42

Signed-off-by: sfc-gh-mvashishtha <mahesh.vashishtha@snowflake.com>
@sfc-gh-mvashishtha sfc-gh-mvashishtha added the NO-CHANGELOG-UPDATES This pull request does not need to update CHANGELOG.md label Sep 30, 2025
@sfc-gh-mvashishtha sfc-gh-mvashishtha marked this pull request as ready for review October 1, 2025 06:40
@sfc-gh-mvashishtha sfc-gh-mvashishtha requested review from a team as code owners October 1, 2025 06:40
@sfc-gh-mvashishtha sfc-gh-mvashishtha marked this pull request as draft October 1, 2025 06:44
@sfc-gh-mvashishtha sfc-gh-mvashishtha marked this pull request as ready for review October 1, 2025 08:07
@sfc-gh-joshi

Copy link
Copy Markdown
Contributor

Is it possible to use pytest -n auto instead in case we change CI runners in the future?

Signed-off-by: sfc-gh-mvashishtha <mahesh.vashishtha@snowflake.com>
@sfc-gh-mvashishtha

Copy link
Copy Markdown
Contributor Author

Is it possible to use pytest -n auto instead in case we change CI runners in the future?

Done.

Signed-off-by: sfc-gh-mvashishtha <mahesh.vashishtha@snowflake.com>
Signed-off-by: sfc-gh-mvashishtha <mahesh.vashishtha@snowflake.com>
Signed-off-by: sfc-gh-mvashishtha <mahesh.vashishtha@snowflake.com>
@sfc-gh-mvashishtha sfc-gh-mvashishtha marked this pull request as draft October 1, 2025 21:02
@sfc-gh-mvashishtha

Copy link
Copy Markdown
Contributor Author

I'm marking this PR as a draft because:

  • it may cause OOMs for AST tests
  • I think we can go even faster with -n 128 since modin tests are network-bound.

@sfc-gh-mvashishtha sfc-gh-mvashishtha changed the title SNOW-2380741: Use all cores in CI. [DRAFT] SNOW-2380741: Use all cores in CI. Oct 1, 2025
@sfc-gh-mvashishtha sfc-gh-mvashishtha changed the title [DRAFT] SNOW-2380741: Use all cores in CI. [DRAFT] SNOW-2380741: Use all logical cores in snowpark tests. Oct 3, 2025
Signed-off-by: sfc-gh-mvashishtha <mahesh.vashishtha@snowflake.com>
@sfc-gh-mvashishtha sfc-gh-mvashishtha marked this pull request as ready for review October 3, 2025 17:11
@sfc-gh-mvashishtha

Copy link
Copy Markdown
Contributor Author

I found that using 128 cores seemed to push the GitHub Actions runners beyond their capacity so that they failed with exit code 143: #3824

I used the previous level of parallelism for AST tests, which seem to use too many resources when we use -n logical.

@sfc-gh-mvashishtha sfc-gh-mvashishtha changed the title [DRAFT] SNOW-2380741: Use all logical cores in snowpark tests. SNOW-2380741: Use all logical cores in snowpark tests. Oct 3, 2025
@sfc-gh-mvashishtha sfc-gh-mvashishtha merged commit ff7b1c2 into main Oct 3, 2025
29 of 31 checks passed
@sfc-gh-mvashishtha sfc-gh-mvashishtha deleted the mavshishtha/SNOW-2380741/use-all-cores-in-ci branch October 3, 2025 20:41
@github-actions github-actions Bot locked and limited conversation to collaborators Oct 3, 2025
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.

Labels

NO-CHANGELOG-UPDATES This pull request does not need to update CHANGELOG.md

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants