Repository navigation
Flake batch 29: rebind diagnostics test retries the non-blocking job snapshot - #1220
Merged
Merged
Conversation
rebind_wait_diagnostics_name_active_job_without_waiting_on_database_slot read the executor's root-job snapshot once. That snapshot is a non-blocking read and reports None while the scheduler holds its lock, so on main (run 41117835406) the diagnostic printed 'root jobs ... None' and missed the held writer. Retry while the database slot stays held, up to 10 s; every assertion is unchanged. Co-authored-by: alfonso-aft <289616620+alfonso-aft@users.noreply.github.com>
Co-authored-by: alfonso-aft <289616620+alfonso-aft@users.noreply.github.com>
Co-authored-by: alfonso-aft <289616620+alfonso-aft@users.noreply.github.com>
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Main's run 41117835406 (commit d3b75e8) failed
subc::tests::rebind_wait_diagnostics_name_active_job_without_waiting_on_database_slot. The diagnostic printedroot jobs (None means scheduler locked): None: the executor's root-job snapshot is a non-blocking read and reports None while the scheduler holds its lock, by design, so a single read can miss the held writer.The test now retries the diagnostic for up to 10 s while it still holds the database slot, until the snapshot gets through. Every assertion is unchanged. It's the same class as the startup-deadline test fixed in #1212.
Test-only. Checked on the build server: the test passed 3 times, formatting and the Windows check passed.
Need help on this PR? Tag
@codesmith-botwith what you need.Summary by cubic
Fixes the flaky
rebind_wait_diagnostics_name_active_job_without_waiting_on_database_slottest.The diagnostic read the executor's root-job snapshot once, and that snapshot reports
Nonewhile the scheduler holds its lock. The test now retries the diagnostic for up to 10 s while the database slot is held, so the assertions check snapshot contents rather than lock timing. All assertions are unchanged; the change is test-only, with a slice descriptor for the flake batch.Written for commit 13ea09e. Summary will update on new commits.