1296057 - Possible race condition when using the bugfiler tool.

Status: RESOLVED FIXED

(Tree Management :: Treeherder, defect)

Description

[Wes Kocher (:KWierso) (Not reading bugmail; email directly if needed)]

I've noticed times when using the bugfiler that the selected failure does not get the classification saved when the new bug is filed. I guess it's some sort of race condition between clicking the submit button, sending off the bug to bugzilla, pinning the failure to the pinboard, classifying the failure with the new bug number, and submitting the classification.


One thing we could do is pre-pin the job the moment the bugfiler button is clicked in the failure summary panel, so we don't need to waste time after the bug is submitted?

Comment 1

[GitHub Autolander Bot]

Attachment: [treeherder] KWierso:filerrace > mozilla:master

Comment 2

[Wes Kocher (:KWierso) (Not reading bugmail; email directly if needed)]

Comment on attachment 8782226 [details] [review]
[treeherder] KWierso:filerrace > mozilla:master

This just moves the call to pin the selected job up to right after the bug filer modal gets opened up. 
I'm hoping this could fix these weird instances where I successfully file a bug, the job gets pinned to the pinboard and the bug number gets assigned to the bug, but it doesn't save the classification for me.

Comment 3

[Phil Ringnalda (:philor)]

May be a different thing, or may be a convenient way to test whether that fixes it, but I get this most consistently when I have two tabs open and file from the second one:

* have inbound open, onlyunstarred
* open the (sig) link from a failure in a new tab, toggle off onlyunstarred
* file from there

Comment 4

[Treeherder GitHub Bugbot]

Commit pushed to master at https://github.com/mozilla/treeherder

https://github.com/mozilla/treeherder/commit/2cb332fa9772c9d14c4c8b31e7d58b916c8a5343
Bug 1296057 - Pin the selected job upon opening bugfiler (#1793) r=emorley

Comment 5

[Wes Kocher (:KWierso) (Not reading bugmail; email directly if needed)]

Lets reopen if this is still happening after this gets deployed.