ci: hunt the wedge with concurrent runs, not CPU spinners (#564) - #567
Merged
Merged
Conversation
The first version of this workflow took 35 arm samples without once reproducing #564, while CI reproduces it about one run in sixteen. Two reasons, both fixed here. It modelled load as four CPU spinners behind a single test binary. CI's `make test` runs four real test binaries at once, so the contention is for disk, memory and the scheduler as well as CPU. Run copies of the fixture concurrently instead: the right kind of load, and four samples per round rather than one. And it stopped on any non-zero exit. The one failure it did catch was ImportReplication -- a different arm flake, the fixture racing its own shutdown and getting "could not connect to server; Connection refused" after the server had logged Shutdown() complete -- which ended the batch at 15 of 20 samples without touching the thing being hunted. Now only a GracefulShutdownUnresponsiveSlave failure stops the batch; everything else is counted and reported per copy. Also distinguishes the two ways the graceful fixture can fail: with a Reap timeout, where #565's interrogation produces the dump this is all for, and without one, which would be a different failure mode worth knowing about.
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
ci: hunt the wedge with concurrent runs, not CPU spinners (#564)
The first version of this workflow took 35 arm samples without once
reproducing #564, while CI reproduces it about one run in sixteen. Two
reasons, both fixed here.
It modelled load as four CPU spinners behind a single test binary. CI's
make testruns four real test binaries at once, so the contention is fordisk, memory and the scheduler as well as CPU. Run copies of the fixture
concurrently instead: the right kind of load, and four samples per round
rather than one.
And it stopped on any non-zero exit. The one failure it did catch was
ImportReplication -- a different arm flake, the fixture racing its own
shutdown and getting "could not connect to server; Connection refused" after
the server had logged Shutdown() complete -- which ended the batch at 15 of
20 samples without touching the thing being hunted. Now only a
GracefulShutdownUnresponsiveSlave failure stops the batch; everything else is
counted and reported per copy.
Also distinguishes the two ways the graceful fixture can fail: with a Reap
timeout, where #565's interrogation produces the dump this is all for, and
without one, which would be a different failure mode worth knowing about.