Actions
action #72085
openflaky/unstable t/05-scheduler-full.t, Failed test 'Allocated maximum number of jobs that could have been allocated'
Status:
Workable
Priority:
Low
Assignee:
-
Category:
Regressions/Crashes
Target version:
Start date:
Due date:
% Done:
0%
Estimated time:
Description
Observation¶
In around 1/400 test runs t/05-scheduler-full fails with
t/05-scheduler-full.t .. 3/?
# Failed test 'Allocated maximum number of jobs that could have been allocated'
# at t/05-scheduler-full.t line 241.
# got: '9'
# expected: '10'
# Looks like you failed 1 test of 1.
Steps to reproduce¶
Follow #59043#note-31
Expected result¶
t/05-scheduler-full.t should be stable, e.g. in 1000/1000 runs.
Suggestions¶
- Reproduce the issue in your local environment, e.g. with
time env runs=400 count_fail_ratio prove -l t/05-scheduler-full.t
using http://github.com/okurz/scripts/bin/count_fail_ratio - Investigate how the check in t/05-scheduler-full.t line 241 can fail, apply fixes and prove stability, e.g. with
time env runs=1000 count_fail_ratio prove -l t/05-scheduler-full.t
Updated by okurz over 4 years ago
- Copied from action #59043: Fix unstable/flaky full-stack test, i.e. remove sleep, and ui tests added
Updated by okurz about 4 years ago
- Copied to action #80800: flaky/unstable t/full-stack.t, Failed test 'test 1 is running', Bailout called. Further testing stopped: URL for os-autoinst cmd srv not available added
Actions