New issue
Advanced search Search tips

Issue 799635 link

Starred by 1 user

Issue metadata

Status: WontFix
Owner: ----
Closed: Jan 9
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: ----
Pri: 3
Type: Bug



Sign in to add a comment

Milo has incorrect step duration

Project Member Reported by martiniss@chromium.org, Jan 5 2018

Issue description

I've seen this across a bunch of builds, but https://ci.chromium.org/buildbot/tryserver.chromium.linux/linux_chromium_asan_rel_ng/520128 is an example.

"Upload to test-results [interactive_ui_tests (with patch)]" takes 11 seconds, as displayed on the above page. However, if you look at the step log itself, in https://logs.chromium.org/v/?s=chromium%2Fbb%2Ftryserver.chromium.linux%2Flinux_chromium_asan_rel_ng%2F520128%2F%2B%2Frecipes%2Fsteps%2FUpload_to_test-results__interactive_ui_tests__with_patch__%2F0%2Fstdout, it takes only 7 seconds internally. This is confirmed by looking at the annotation stream (I couldn't figure out how to link this, but `logdog cat -timestamps utc chromium/bb/tryserver.chromium.linux/linux_chromium_asan_rel_ng/520128/+/recipes/annotations` seems to work for me). According to that, the step takes 7 seconds. In addition, the step before the upload ends a few milliseconds before the upload step starts, and the step after starts a few milliseconds after the upload ends.

We seem to be apparently losing 4 seconds somewhere. Or, milo is using wrong data. Either way, something seems wrong here.
 
Components: -Infra>Platform>Milo Infra>Platform>Milo>Buildbot
Milo uses buildbot data for buildbot builds.  So this would be a buildbot bug.

https://screenshot.googleplex.com/M2o7SieG6Gw

Comment 2 by efoo@chromium.org, Jan 6 2018

How important is the fix? Considering the Buildbot workflow in Milo will be deprecated in the near future, I'm not sure we should spend time on this unless it's absolutely necessary or if migration somehow gets drags out. 
P3 is fine. If the migration drags on, I'd like for this to be fixed. This bug is more for me to realize where these discrepancies come from. Leaving it as is is fine.

Comment 4 by efoo@chromium.org, Jan 9 2018

Labels: LUCI-Afterglow
Marked for Afterglow for review. 
Project Member

Comment 5 by sheriffbot@chromium.org, Jan 9

Status: Archived (was: Unconfirmed)
Issue has not been modified or commented on in the last 365 days, please re-open or file a new bug if this is still an issue.

For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot
Status: WontFix (was: Archived)
wontfix because buildbot

Sign in to add a comment