action #15310
closed[functional][u][easy] wait_still_screen in start_x11_program wasting time when program with blinking cursor is started
0%
Description
observation¶
See
https://openqa.opensuse.org/tests/315091#step/kate/13
It's already four minutes after
https://openqa.opensuse.org/tests/315091#step/kate/11
problem¶
Wasting a lot of time waiting for still screen which is never happening as the cursor keeps blinking
suggestion¶
- Make
wait_still_screenfails fatal, why not let the test fail so that we see these problems? Or at least some warning in the log - Tweak start_x11_program to not wait in these specific cases
Further details¶
Updated by okurz about 6 years ago
- Subject changed from wait_still_screen in start_x11_program wasting time when program with blinking cursor is started to [functional]wait_still_screen in start_x11_program wasting time when program with blinking cursor is started
- Target version set to Milestone 16
Updated by okurz almost 6 years ago
- Subject changed from [functional]wait_still_screen in start_x11_program wasting time when program with blinking cursor is started to [functional][u] wait_still_screen in start_x11_program wasting time when program with blinking cursor is started
- Due date set to 2018-06-05
Updated by okurz almost 6 years ago
- Subject changed from [functional][u] wait_still_screen in start_x11_program wasting time when program with blinking cursor is started to [functional][u][easy] wait_still_screen in start_x11_program wasting time when program with blinking cursor is started
- Status changed from New to Workable
- Difficulty set to easy
Updated by mgriessmeier almost 6 years ago
- Due date changed from 2018-06-05 to 2018-06-19
Updated by okurz almost 6 years ago
- Due date set to 2018-07-17
- Target version changed from Milestone 16 to Milestone 17
Updated by okurz almost 6 years ago
- Target version changed from Milestone 17 to Milestone 17
Updated by okurz almost 6 years ago
- Target version changed from Milestone 17 to Milestone 18
Updated by okurz almost 6 years ago
- Due date changed from 2018-07-17 to 2018-07-31
It's hackweek time!
Updated by okurz almost 6 years ago
- Status changed from Workable to In Progress
- Assignee set to okurz
Looking into start_x11_program now so I can take a look into this as well.
Updated by okurz over 5 years ago
- Status changed from In Progress to Workable
- Assignee deleted (
okurz)
So worked on start_x11_program lately and I am finished with my part. The original issue is still true, e.g. see in https://openqa.opensuse.org/tests/713517/file/autoinst-log.txt
[2018-07-26T08:18:48.0453 CEST] [debug] /var/lib/openqa/cache/openqa1-opensuse/tests/opensuse/tests/x11/kate.pm:22 called testapi::x11_start_program
[2018-07-26T08:18:48.0453 CEST] [debug] <<< testapi::wait_still_screen(stilltime=3, similarity_level=47, timeout=30)
[2018-07-26T08:19:18.0288 CEST] [debug] >>> testapi::wait_still_screen: wait_still_screen timed out after 30
[2018-07-26T08:19:18.0288 CEST] [debug] /var/lib/openqa/cache/openqa1-opensuse/tests/opensuse/tests/x11/kate.pm:22 called testapi::x11_start_program
that is 30s wasted. Could be solved by either getting rid of the wait_still_screen or tweaking the similarity_level parameter.
Updated by jorauch over 5 years ago
- Status changed from Workable to In Progress
Actually working on this since friday.
Updated by jorauch over 5 years ago
Since the fail appeared with similarity level 47 I would suppose to set the level to 45 and check if this has unwanted side effects
Updated by mgriessmeier over 5 years ago
- Due date changed from 2018-07-31 to 2018-08-14
Updated by jorauch over 5 years ago
Created PR:
https://github.com/os-autoinst/os-autoinst-distri-opensuse/pull/5500
This should avoid wasting time by waiting for blinking cursors
Next I will try to get wait_still_screen to tell us when it timed out so we see these situations better
Updated by jorauch over 5 years ago
Regarding the first suggestion we should not change wait_still_screen itself but try to check the return values when it is really critical
Also we have Log Entries so not sure, how to continue here
Updated by okurz over 5 years ago
PR merged, great! Yes, you are right in all your points. We could go with changing "wait_still_screen" to "assert_still_screen" by default and only use "wait_still_screen" or a hypothetical "check_still_screen" only in cases we want to explicitly check the return value. I already tried this with a new function assert_screen_change which is part of os-autoinst already doing the equivalent for wait_screen_change but I think it does not even work properly. When you want you can simply close the ticket or create new feature requests on the openQA issue tracker to have assert_screen_change and assert_still_screen functions with proper unit tests in os-autoinst. As a ticket first but if you are motivated we can also bring it into our backlog and work on that.
Updated by jorauch over 5 years ago
- Related to action #39116: Add assert_still_screen functionality added
Updated by jorauch over 5 years ago
Created https://progress.opensuse.org/issues/39116 as a followup
Will add an output of the real similarity level to the logs and then close this
Updated by jorauch over 5 years ago
To print the last similarity level I would have to change the scope of $sim out of the loop
Updated by jorauch over 5 years ago
- Status changed from In Progress to Feedback
After an adventurous journey for testing this I created https://github.com/os-autoinst/os-autoinst/pull/1009 to print the similarity level to the log
Updated by okurz over 5 years ago
- Due date changed from 2018-08-14 to 2018-08-28
bulk move to next sprint as could not be discussed in SR
Updated by jorauch over 5 years ago
- Status changed from Feedback to Resolved
PR is merged, therefore closing this ticket