Added by maritawerner about 8 years ago. Updated almost 8 years ago.
0%
Description
After the installation the pool channels should be added and tested but it seems that the testcase zypper_lr does not do that properly, no needles are visible,
e.g. https://openqa.suse.de/tests/356456/modules/zypper_lr/steps/1
So, I guess the task is to make sure that zypper_lr verifies that zypper lr produces something sensible apart from verifying the command exits with return code 0.
I think the idea Marita had referencing the specific example is that the "zypper_lr" test case does not check that the sdk ftp repo is actually present. I fear needling the expected output will be quite error prone and tedious to support all different cases we have. I am thinking of wait_serial checking for what you expect, which might be tricky as we would need to know the repo paths for the corresponding addon. Can you investigate how the output of zypper_lr should look like per addon?
Also attending https://progress.opensuse.org/projects/openqatests/wiki#Definition-of-READY-for-new-tests might help as a checklist to get it covered. Ask marita for details if you are unsure.
Here
https://wiki.microfocus.net/index.php/SLE12_SP2_Channels_Checking_Table
you can find a Table of all Channels. But I think that update channels are only visible with real SCC and not with proxy SCC. I have to double check that next week.
Did not get you messages as I failed to assign this PR to me. But I believe I came up with a similar approach: https://github.com/os-autoinst/os-autoinst-distri-opensuse/pull/1495. To test all those addons combinations I extended addons_product_sle.pm: https://github.com/os-autoinst/os-autoinst-distri-opensuse/pull/1496. Plus quite a few needles: https://gitlab.suse.de/openqa/os-autoinst-needles-sles/merge_requests/160/.
Looking at Acceptance test groups at openqa.s.de the test is being executed and is not failing anywhere, I believe this PR should be closed as resolved.
yes, I agree. Feel free to set the status to RESOLVED yourself next time, especially if you backed your decision by hard facts, i.e. as you stated, no failures in recent builds :-)