Closed
Bug 1173882
Opened 10 years ago
Closed 10 years ago
Autophone - should use updated attempts when inserting a new job due to a retry
Categories
(Testing Graveyard :: Autophone, defect)
Testing Graveyard
Autophone
Tracking
(firefox41 affected)
RESOLVED
FIXED
Tracking | Status | |
---|---|---|
firefox41 | --- | affected |
People
(Reporter: bc, Assigned: bc)
References
Details
(Keywords: regression)
Attachments
(2 files)
2.36 KB,
patch
|
gbrown
:
review+
|
Details | Diff | Splinter Review |
879 bytes,
patch
|
gbrown
:
review+
|
Details | Diff | Splinter Review |
Last night I noticed a job which continually submitted retries:
https://treeherder.mozilla.org/#/jobs?repo=try&exclusion_profile=false&filter-searchStr=autophone&revision=6e083a56ff88
I saw this again locally when testing --override-build-dir where incompatible api builds were being served to builds. The job would be retried and the newly inserted job would have attempts initialized to 0 which meant that retried jobs would not be purged after exceeding the maximum number of attempts.
see https://github.com/mozilla/autophone/blob/master/worker.py#L629
The fix is to change Jobs.new_job to take an attempts keyword argument with default 0 and to use that to increment the attempts for retried jobs.
Assignee | ||
Comment 1•10 years ago
|
||
Attachment #8621243 -
Flags: review?(gbrown)
![]() |
||
Updated•10 years ago
|
Attachment #8621243 -
Flags: review?(gbrown) → review+
Assignee | ||
Comment 2•10 years ago
|
||
Status: ASSIGNED → RESOLVED
Closed: 10 years ago
Resolution: --- → FIXED
Assignee | ||
Comment 3•10 years ago
|
||
I shouldn't have incremented the attempts in the new retried job. Just passing the existing attempts would have been correct.
Status: RESOLVED → REOPENED
Resolution: FIXED → ---
Assignee | ||
Comment 4•10 years ago
|
||
Attachment #8621729 -
Flags: review?(gbrown)
![]() |
||
Comment 5•10 years ago
|
||
Comment on attachment 8621729 [details] [diff] [review]
bug-1173882-followup.patch
Review of attachment 8621729 [details] [diff] [review]:
-----------------------------------------------------------------
Sorry, I should have flagged that.
Attachment #8621729 -
Flags: review?(gbrown) → review+
Assignee | ||
Comment 6•10 years ago
|
||
Status: REOPENED → RESOLVED
Closed: 10 years ago → 10 years ago
Resolution: --- → FIXED
Assignee | ||
Comment 7•10 years ago
|
||
fwiw, this was a regression from bug 1165403. Prior to that check in, the new job generated from the retry would be attached to the existing row in the jobs table. After the check in, a new row would be generated in the jobs table for the build if duplicates were allowed. This new row would have had attempts 0 initially which was the cause of our problem.
Blocks: 1165403
Keywords: regression
Updated•3 years ago
|
Product: Testing → Testing Graveyard
You need to log in
before you can comment on or make changes to this bug.
Description
•