ReleaseEngineering/BuildFaster/Meetings/2011-07-20
From MozillaWiki
Contents
Details
July 20, 2011 @ 9am PDT / 12pm EDT
Mountain View room GIGO dial-in x92 95272 (vidyo)
Backchannel #buildfaster on irc.mozilla.org
Agenda
Action items from last time
- catlee to file bug on getting metrics on pieces of end-to-end time. DONE bug 669930
- ctalbert to follow up on slow tests DONE bug 670229
- ctalbert to file bug on getting commands to reproduce test runs on developer machine (give URLs to packages to download + steps to run) DONE bug 673001
- ted to file a bug on investigating rsyncing of test package contents DONE bug 669952
- catlee to get IT bug filed for minidump processing DONE bug 669947
- ctalbert to follow up on sdwilsh'es tree status DONE
- ted to file a bug on enabling optimize on debug builds DONE bug 669953
- ted to ask dbaron about tracemalloc on debug builds, is it a perf hit by default?
Review [buildfaster:p1] bugs
- bug 367393
- Still picking out parts of packed.js while rewriting some of the useful functions.
- Finding which tests have dependencies on packed.js and instead depend on the respective parts of MochiKit.
- bug 479352 - backed out, ctalbert to help figure out what the issue is with m1 tests
- bug 561754 - waiting on the server to be set up for sending the symbol processing to. Re-casting the related talos bug so that it's clear what the solution is there and moving it to talos component as talos work is required to enable this functionality.
- Blocked - Waiting for [?] to show up. Moving to talos component.
- bug 623617
- Has working prototype set up. Wiki pages will be set up regarding changes
- bug 658313
- Finishing up Talos changes to allow PGO builds. Should be ready for review shortly. Currently working for mozilla-central
- Worried about slave capacity from running the extra every-4-hour PGO builds
- Ties up one slave per platform per branch per day
- Can we run PGO only on nightly builds instead?
- If we do only nightly PGO builds, maybe use Try for bisecting if we hit a failure?
- Should make sure that TBPL distinguishes PGO vs. non-PGO builds on the same changeset to avoid confusion and notice PGO-specific test failures
- bug 659328
- Working on a patch that needs some refactoring, should be up for next week
- bug 669930
- samliu is working on getting data dumps and displaying metrics properly
- bug 669988
- motivation was faster build times from doing single-arch builds; we're getting new build hardware soon, so this may not be an issue. -> p5
- bug 670229
- Need to distinguish between tests that are slow (>10s) and tests that are supposed to take a longer time
- Someone should look at each test and see if there's no obvious reason for it being slow, file bugs on those
- See if certain tests are much slower on debug, hard to give a good heuristic
Review current activities
- See list of current activities
- Is everything up to date?
- Tag bugs with [buildfaster:p1], [buildfaster:p2], etc.
- What do we do with jgriffin's test analysis?
Review [buildfaster:?] bugs
Questions
From last time:
- run 10.5 debug builds on 10.6 machines? It increases capacity of 10.6 pool. Low low priority. Loosely related.
- shall we break up the current activities section into bug being worked on and bugs in the back burner?
Actions for next time
- bug 670229 - follow up bugs for individual tests that are slow
- Ted to follow up on tracemalloc question -- filed bug 673198
- go through p2 bugs and see if we need to escalate
