QA » openQA Project » openQA Tests

This is an autogenerated message for openQA integration by the openqa_review script:

This bug is still referenced in a failing openQA test: minimal_x
https://openqa.suse.de/tests/5669331

To prevent further reminder comments one of the following options should be followed:

1. The test scenario is fixed by applying the bug fix to the tested product or the test is adjusted
2. The openQA job group is moved to "Released"
3. The label in the openQA scenario is removed

Running some statistical investigation (https://progress.opensuse.org/projects/openqatests/wiki/Wiki#Statistical-investigation) to see if some any issue can be spotted:

for i in {003..100} ; do \
/usr/share/openqa/script/clone_job.pl \
--from https://openqa.suse.de \
--host https://openqa.suse.de \
-v  6357721       \
--skip-chained-deps \
ISO= \
SCHEDULE=tests/boot/boot_to_desktop,tests/console/cleanup_qam_testrepos,tests/console/ncurses \
_SKIP_POST_FAIL_HOOKS=1 \
QAM_TESTREPO_FAIL=purge \
TEST=mgrifalconi-investigation-ncurses-$i \
BUILD=mgrifalconi-investigation-ncurses \
_GROUP="mgrifalconi-investigation-ncurses" \
 ; done

Using mau-extratest as blueprint to copy, since ncurses is currently tested only on installation test suites, and they take a lot of time.

Observation: Why don't we move out as much as possible tests from installation jobs, to reduce risk of breaking the job and all its dependencies?
Does ncurses in particular need to be tested right after the installation? Can't be moved to mau-extratest?

Results: https://openqa.suse.de/tests/overview?build=mgrifalconi-investigation-ncurses

mgrifalconi wrote:

Observation: Why don't we move out as much as possible tests from installation jobs, to reduce risk of breaking the job and all its dependencies?
Does ncurses in particular need to be tested right after the installation? Can't be moved to mau-extratest?

I think QE-Core already has it on the backlog to move more and more tests into "extratests" which boot from existing disk images rather than doing all the tests within the same installation based scenario. However the approach of loading from images is not implemented. For these cases there can be "directly chained dependencies" to ensure that after an installation on a specific (physical) machine the installed OS is booted and subsequent tests can be executed.

We agreed to move such test out of installation jobs to extratests or other jobs that boot from a disk, since there is no reason to run it right after installation.

It seems not possible to run the dialog command as part of a shell script, since it blocks there and the script never ends (waiting for a dialog). Considering it's only 2 commands, I propose to skip the switch to bash.

About changing the logic of zypper_call, wouldn't make sense to have a separate progress issue and try to advertise it a bit since everyone is using such method?

Created https://progress.opensuse.org/issues/95362 about zypper_call refactoring

Created https://progress.opensuse.org/issues/95365 about moving ncurses out of installation tests

Will not move it to bash script since it's only 2 lines of code to be moved.

PR for using serial terminal on zypper_call https://github.com/os-autoinst/os-autoinst-distri-opensuse/pull/12880