Project

General

Profile

Actions

action #153610

open

[sporadic] Unstable OBS checks in devel:openQA/os-autoinst on aarch64

Added by okurz 11 months ago. Updated 11 months ago.

Status:
New
Priority:
Low
Assignee:
-
Category:
Regressions/Crashes
Target version:
QA (public, currently private due to #173521) - future
Start date:
2024-01-16
Due date:
% Done:

0%

Estimated time:

Description

Observation

Over the past days I have observed that https://build.opensuse.org/package/live_build_log/devel:openQA/os-autoinst/openSUSE_Factory_ARM/aarch64 sometimes shows failed checks. As this URL only shows the latest result sometimes I received an email about problems with not enough details but then following the URL I see "Build succeeded" so I have to assume a sporadic issue.

Today I see

[  277s] 3: [08:20:20] ./t/28-signalblocker.t .................................. 
[  277s] 3: ok 1 - initially one thread
[  277s] 3: # threads used: 3 of 3
[  277s] 3: ok 2 - at least 3 threads created
[  277s] 3: # received SIGCHLD 1
[  277s] 3: # received SIGCHLD 2
[  277s] 3: not ok 3 - no new threads after searching for a needle
[  277s] 3: 
[  277s] 3: #   Failed test 'no new threads after searching for a needle'
[  277s] 3: #   at ./t/28-signalblocker.t line 61.
[  277s] 3: #     '4'
[  277s] 3: #         <=
[  277s] 3: #     '3'

checking reproducibility locally with runs=1000 count_fail_ratio prove -I. t/28-signalblocker.t using https://github.com/okurz/scripts/blob/master/count_fail_ratio shows no clearly reproducible issue as of now:

## count_fail_ratio: Run: 1000. Fails: 0. Fail ratio 0±0%. No fails, computed failure probability < .30%

I suggest to wait for further results from OBS, check if the checks fail in the same step and only on aarch64 and then we know if it's only reproducible in that special scenario.

Actions

Also available in: Atom PDF