-
Notifications
You must be signed in to change notification settings - Fork 1.2k
Change _TestHelixAgentPool from 19H1 to RS5 #975
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Conversation
Hello @vatsan-madhavan! Because this pull request has the Do note that I've been instructed to only help merge pull requests of this repository that have been opened for at least 30 minutes. No worries though, I will be back when the time is right! 😉 p.s. you can customize the way I help with merging this pull request, such as holding this pull request until a specific person approves. Simply @mention me (
|
@msftbot, require at least 1 sign off. |
@msftbot, require approvals from any one of the following: @vatsan-madhavan, @rladuca, @SamBent, @stevenbrix, @ryalanms |
This time RS5 seemed to complete quickly, but generic Win10 and Win7 took a lot of time. I’m not sure what the right solution should be. @MattGal any suggestions? |
@vatsan-madhavan I'm observing that certain teams are doing work items that take 1.5+ hours EACH due to setting special environment variables for JIT stress and such; when these teams (i.e. coreCLR) run their tests on any given queue, that queue will be backed up a bit because of the slow running work items sent. I'll try to identify the folks doing this and see if we can maybe not do this every PR, but for now the best solution is to have longer build timeouts; if you don't the work still gets done, but the build gets retried; this is only making the problem worse and not improving your speeds at all. |
…s is the only one that seems immune to timeouts - Limit Helix test runs to the open build pipline. Do not repeat the same tests in the internal pipeline.
Alright, we'll try to up our timeout to 120 mins for now. Is there a tracking issue we can follow to keep up? I'd really like to keep our build times not become arbitrarily long. Can we chat about the possibility of getting separate Helix queues for our team (perhaps shared with WinForms) maybe? /cc @AdamYoblick |
a7e35a7
to
d035cf5
Compare
We will add back Win7 after Preview 7. For now, we will test RS5 only. |
@msftbot, require at least 1 sign off. |
@msftbot, require approvals from any one of the following: @vatsan-madhavan, @rladuca, @SamBent, @stevenbrix, @ryalanms |
Changing
_TestHelixAgentPool
from 19H1 to RS5 seems to bring down execution time in Helix to approx 15 mins.https://dev.azure.com/dnceng/public/_build/results?buildId=226613&view=logs
Looks like the 19H1 pool may have been the reason our builds have been suffering from chronic timeouts, much like the ServerRS5 pool prior to that.
Fixes #952