Actions
action #43031
closedFix failing t/ui/13-admin.t
Description
Seems like t/ui/13-admin.t fails now quite often on Travis:
# Failed test 'advanced_kde not selectable twice'
# at ./t/ui/13-admin.t line 416.
# Structures begin differing at:
# $got->[1] = 'advanced_kde'
# $expected->[1] = 'client1'
# Looks like you failed 1 test of 8.
# Failed test 'edit mediums'
# at ./t/ui/13-admin.t line 492.
Use of uninitialized value $expr in concatenation (.) or string at /usr/lib/perl5/5.18.2/B/Deparse.pm line 1370.
# Looks like you failed 1 test of 15.
Use of uninitialized value $expr in concatenation (.) or string at /usr/lib/perl5/5.18.2/B/Deparse.pm line 1370.
./t/ui/13-admin.t ....................
Dubious, test returned 1 (wstat 256, 0x100)
Failed 1/15 subtests
notes¶
- I am not able to reproduce the issue locally (not by running the test directly, but also not by running it within Docker)
- affected PRs: https://github.com/os-autoinst/openQA/pull/1843 (only changes documentation), https://github.com/os-autoinst/openQA/pull/1844
Updated by EDiGiacinto about 6 years ago
Also https://github.com/os-autoinst/openQA/pull/1783 is affected by it (after rebasing)
Updated by mkittler about 6 years ago
- Status changed from New to In Progress
There is no AJAX involved in the relevant part of the JavaScript code. Adding a wait_for_ajax nervelessness or even a regular sleep doesn't work. So I still have no clue why the JavaScript behaves wrong when the tests are executed within Travis. But it seems like it is now almost reproducible 100 % of the time (within Travis).
If I don't get a clue soon, we could also temporarily make that test non-fatal. Better than having red status on each PR.
Updated by mkittler about 6 years ago
- Status changed from In Progress to Resolved
Updated by coolo about 6 years ago
- Target version changed from Current Sprint to Done
Actions