Repository navigation
ci: raise default A/B max iterations from 4 to 16 - #6373
Merged
Merged
Conversation
The A/B pipeline fails on noise: build firecracker-microvm#1095 compared two revisions with byte-identical binaries and still failed three jobs. ping_latency run means are bimodal on several hosts (e.g. 38/56 us on m7g), and since the permutation test treats the 500 pings of a run as independent samples, one run in the other mode is enough to trip the 5% threshold. Replaying ab_test.py on builds 1080-1095 gives a 1.3% per-job false positive rate at 4 iterations (43-77% odds of a clean build) and 0.01-0.1% at 16 (94-99%). Only failing cases are re-run, so passing jobs cost little extra. Signed-off-by: Marco Marangoni <mamarang@amazon.com>
marco-marangoni
added
the
Status: Awaiting review
Codecov Report✅ All modified and coverable lines are covered by tests. Additional details and impacted files@@ Coverage Diff @@
## main #6373 +/- ##
=======================================
Coverage 83.59% 83.59%
=======================================
Files 282 282
Lines 33606 33606
=======================================
Hits 28094 28094
Misses 5512 5512
Flags with carried forward coverage won't be shown. Click here to find out more. ☔ View full report in Codecov by Harness. 🚀 New features to boost your workflow:
|
JamesC1305
approved these changes
Nov 3, 2026
zulinx86
approved these changes
Nov 3, 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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Changes
Increase A/B test iterations.
I've also increase the timeout on BuildKite from 150m to 360m. We recently reduced the number of test cases we run on A/B test, so I hope this iteration increase isn't that impactful.
I chose the number 16 because I simulated the current numbers of the network test (the one with the highest false positive rate), and with 16 iteration the probability of one test case failing 16 times in a row goes basically to 0.01% per job.
Reason
Reduce false positive: each iteration collects more data for tests that fail, so iterating more increases confidence in the result.
In theory true positive rate isn't impacted, because datapoints are remembered across iterations.
License Acceptance
By submitting this pull request, I confirm that my contribution is made under
the terms of the Apache 2.0 license. For more information on following Developer
Certificate of Origin and signing off your commits, please check
CONTRIBUTING.md.PR Checklist
tools/devtool checkbuild --allto verify that the PR passesbuild checks on all supported architectures.
tools/devtool checkstyle --no-clippyto verify that the PRpasses the automated style checks.
how they are solving the problem in a clear and encompassing way.
in the PR.
CHANGELOG.md.Runbook for Firecracker API changes.
integration tests.
TODO.rust-vmm.