Linux tsan crashtest tasks often lost, task resolved with 'CLAIM_EXPIRED'
Categories
(Core :: Sanitizers, defect)
Tracking
()
| Tracking | Status | |
|---|---|---|
| firefox-esr78 | --- | unaffected |
| firefox89 | --- | unaffected |
| firefox90 | --- | unaffected |
| firefox91 | --- | fixed |
People
(Reporter: aryx, Assigned: jonco)
References
(Blocks 1 open bug, Regression)
Details
(Keywords: regression)
Attachments
(1 file)
During the last week, the Linux tsan crashtest task has failed 12 times (= 6 attempts on the same push, higher failure rate if one looks at the attempts) because the worker stopped communicating with taskcluster (e.g. it crashed) - first failure.
Jon, could this be from bug 1715562?
|
Assignee | |
Comment 1•4 years ago
|
||
(In reply to Sebastian Hengst [:aryx] (needinfo on intermittent or backout) from comment #0)
It could be. Is it still happening since bug 1716940 landed?
|
Reporter | |
Comment 2•4 years ago
|
||
The issue still affects the latest pushes and backfills indicate it started with bug 1716940.
|
||
Updated•4 years ago
|
|
||
Updated•4 years ago
|
|
|
Assignee | |
Comment 3•4 years ago
|
||
I can run these tests locally without issue (modulo some TSAN warnings) but they do take a lot of memory (~10GB for the first chunk). I don't think this is directly related to my push (note the previous push also has some failures).
Increasing the chunking fixes the problem on try.
|
|
Assignee | |
Comment 4•4 years ago
|
||
This fixes test crashes suspected to be caused by memory exhaustion.
|
||
Comment 6•4 years ago
|
||
| bugherder | ||
| Comment hidden (Intermittent Failures Robot) |
|
||
Updated•4 years ago
|
Description
•