action #37628

Flaky scheduler test: 'Simulation of heavy unstable load' at t/05-scheduler-full.t

Added by szarate almost 2 years ago. Updated 7 months ago.

Status:ResolvedStart date:21/06/2018
Priority:NormalDue date:
Assignee:okurz% Done:

0%

Category:Concrete Bugs
Target version:QA - future
Difficulty:
Duration:

Description

Sometimes we get that scheduler-full test failing:

ok 38 - Scheduler clock got reset
# Running scheduler step

$VAR1 = [];

ok 39 - timeout in range 1000->3000 (setted tick 2000, real tick occurred at 2002)
ok 40 - Tick is at the expected value (4000) (failures 1) (no_actions 1)
ok 41 - All failed allocation on second step - workers were killed
not ok 42 - Failure count(1) should be >=8, since we took too much time to schedule
#   Failed test 'Failure count(1) should be >=8, since we took too much time to schedule'
#   at t/05-scheduler-full.t line 261.
ok 43 - Tick is at the expected value

History

#1 Updated by okurz 10 months ago

  • Category set to Concrete Bugs

#2 Updated by okurz 8 months ago

  • Status changed from New to Resolved
  • Assignee set to okurz

I do not recall seeing this while tracking flaky tests. Not feeling very confident but 577a921a1 with a remove of d-bus or 535f87443 , a bigger restructuring, could have fixed this.

#3 Updated by okurz 8 months ago

https://github.com/os-autoinst/openQA/pull/2259 to remove "travis_retry" again after test stabilization – and proper testing in the PR

Also available in: Atom PDF