clear console seems to not work everytime
|Category:||Bugs in existing tests|
|Target version:||openQA Project - Milestone 1|
clear console seems to not work, e.g. see
- cryptlvm_minimal_x: https://openqa.suse.de/tests/175735 (reference https://openqa.suse.de/tests/168844)
- logfile: https://openqa.suse.de/tests/175735/file/autoinst-log.txt: "needle(s) 'cleared-console' not found"
#2 Updated by RBrownSUSE about 4 years ago
- Assignee set to RBrownSUSE
- % Done changed from 0 to 80
#5 Updated by RBrownSUSE about 4 years ago
- Assignee changed from RBrownSUSE to okurz
Proposed fix superseded by https://github.com/os-autoinst/os-autoinst-distri-opensuse/pull/926
Assigning to Oli
#8 Updated by okurz about 4 years ago
My PR got merged but I don't understand what is happening here:
>>> assert_screen: match=broken-console-font timed out after 5 <<< type_string(string='clear ', max_interval=250) <<< assert_screen(mustmatch='cleared-console', timeout=30) MATCH(consoletest_setup-20141125:0.96) MATCH(cleared-console-autoyast-font-20141126:0.96) no match 30 MATCH(consoletest_setup-20141125:1.00) MATCH(cleared-console-autoyast-font-20141126:1.00) >>> assert_screen: match=cleared-console timed out after 30 <<< mydie(cause_of_death=[ 'needle(s) \'cleared-console\' not found' ])
After the call to "clear" there is
no match 30
which means that there was no match but we still have a timeout of "30 tries". correct?
why don't I see a decrease to 0 before failing then?
#9 Updated by okurz about 4 years ago
ok, I think I understood it better.
no match <high_number>could mean it exited early from the loop because the screen did not change at all --> check in source code and try to improve log message
no match 2and then fail could mean an "off-by-one" incorrectness, return from the loop in the second-to-last run not the last run --> crosscheck and potentially improve