[Test] Fix issues causing intermittent test failures: test_efa, test_ad_integration, test_queue_parameters_update, test_slurm, test_multiple_efs#7357
Merged
gmarciani merged 8 commits intoApr 30, 2026
Conversation
gmarciani
changed the title
gmarciani
added
skip-changelog-update
himani2411
reviewed
Apr 30, 2026
gmarciani
force-pushed
the
wip/mgiacomo/3160/fix-tests-0429-1
branch
from
1742f7b to
dd37b98
Compare
himani2411
previously approved these changes
Apr 30, 2026
…large on eu-north-1 to reduce the risk of insufficient capacity exceptions.
…heck Add a sleep before the ls command inside the switch-user session to allow the PAM hook time to finish generating the user's SSH key. Without it, ls may run before key generation completes, causing sporadic "Permission denied" failures.
…s EFS data from AL2 to AL2023. AL2 is reaching end of life in June 2026. We also believe that moving to AL2023 could benefit from a more robust networking driver bootstrap, preventing the sporadic networking failures observed in this test.
…sary dependency on NFS shared storage. In particular, we replaced the shared-file round-trip with an `srun` that streams output back right away. This removes a flaky dependency on NFS cross-client visibility, which is already covered by storage-specific tests and is just noise in this test.
…initely. The test could hang indefinitely when dumping job output due to a weak condition and the lack of SSH timeout. When this happens, it could also have impact on other tests. In fact, if the test hangs for too long, our watchdog mechanism would kill all the tests that are sharing the same VPC with this one as a way to protect from infinite test executions.
gmarciani
force-pushed
the
wip/mgiacomo/3160/fix-tests-0429-1
branch
from
51bdbf4 to
a9495d7
Compare
This is required to prevent test failures with pytest 8+, which is pulled in when upgrading to Python 3.10+.
gmarciani
force-pushed
the
wip/mgiacomo/3160/fix-tests-0429-1
branch
from
a9495d7 to
7a28d8c
Compare
himani2411
approved these changes
Apr 30, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Description of changes
Fixed issues causing intermittent failures on the following tests:
test_efa: Replaced hpc5a.48xlarge on us-east-2 with hpc6a.48xlarge on eu-north-1 to reduce the risk of insufficient capacity exceptions.test_ad_integration: Fixed a race condition in the SSH key generation check by adding a sleep before the ls command inside the switch-user session, giving the PAM hook time to finish generating the user's SSH key and avoiding sporadic "Permission denied" failures.test_queue_parameters_update: Prevented the test from hanging indefinitely when dumping job output, caused by a weak condition and the lack of an SSH timeout. Hangs in this test could also trigger the watchdog to kill unrelated tests sharing the same VPC.test_multiple_efs: Switched the auxiliary instance that writes EFS data from AL2 to AL2023. AL2 reaches end of life in June 2026, and AL2023 should provide a more robust networking driver bootstrap, preventing the sporadic networking failures observed in this test.test_slurm: Removed an unnecessary dependency on NFS shared storage by replacing the shared-file round-trip with an srun that streams output back directly. NFS cross-client visibility is already covered by storage-specific tests and was just noise here.Also, included the following improvements:
Tests
test_efatest_ad_integrationtest_queue_parameters_updatetest_slurmtest_multiple_efsBy submitting this pull request, I confirm that my contribution is made under the terms of the Apache 2.0 license.