BotUpdateSlow should provide more guidelines in playbook |
|||||
Issue descriptionIt's not clear what is "slow" and what is not. In particular, the only guideline seems to be that bot_update should be under 600 seconds and it's not clear what is the expected normal time for updating individual repos, e.g. it's not clear whether 4 minutes for gclient-syncing v8 is a lot or normal. Please add respective guidelines into the playbook.
,
Mar 23 2017
Moved to Infra>CQ since this is a service specific alert update.
,
Mar 23 2017
Hmm, IDK if it's really CQ specific? If we were going to assign it to a service I think Buildbot might be more accurate?
,
Mar 23 2017
Yeah, it's 100% not CQ specific.
,
Apr 4 2017
@sergiyb -- have you seen the trooper playbook entry: https://chrome-internal.googlesource.com/infra/infra_internal/+/master/doc/troopers/playbook.md#BotUpdateSlow? I can add some notes about finding the step logs for the specified builder. Do you think that would help? I don't have anything to add beyond that. I know it was frustrating to receive this alert in the past. Now that the bucketer has been fixed, if this alert fires than there really is something wrong, or there is a regression in bot_udpate performance.
,
Apr 4 2017
Assigning to sergiyb for a response.
,
Apr 5 2017
Agreed, now that the bucketer is fixed it should not be an issue. Let's close this as WontFix. |
|||||
►
Sign in to add a comment |
|||||
Comment 1 by dsansome@chromium.org
, Mar 15 2017Status: Assigned (was: Untriaged)