QA » openQA Project » openQA Tests » qe-yam

you need more than one run to probe that the RAM make it pass, it could be any other temporary thingy. For example today all the failures are gone.
I would recommend next time statistical analysis, at least 10 for example.
On the other hand with a gnome most likely we need that amount of RAM by default.

We notice similar issue https://openqa.suse.de/tests/13960449#step/grub_test/2 and QEMURAM=4096 didn't help.

It seems issue for INSTALLER_SELF_UPDATE, if set it as null then passed: http://openqa.suse.de/tests/13960826#

too many self-updates can eat the memory but he problem is normally seen on booting.
@leli could you apply the statistical analysis I mentioned above and if the self_update is what really makes the difference, communicate it in the proper Slack channel for maintainers of those channels?

we are also facing the same issue. Weirdly enough, the desktop installation with 2GB passes, while the minimal installation with 4GB fails.

JERiveraMoya wrote in #note-5:

too many self-updates can eat the memory but he problem is normally seen on booting.
@leli could you apply the statistical analysis I mentioned above and if the self_update is what really makes the difference, communicate it in the proper Slack channel for maintainers of those channels?

Asked this in team-lsg-qe-openqa-review channel: https://suse.slack.com/archives/C02D16TCP99/p1712560469995439

We had another strange failure in installation https://openqa.suse.de/tests/13963317#step/await_install/77 and looks that clearing INSTALLER_SELF_UPDATE helped as well https://openqa.suse.de/tests/13964002#step/logs_from_installation_system/28.

• After declining 33123, it seems no such failures anymore, all MU jobs for building 20240408 passed.

Yes, I confirm the same. There are no installer issues in HPC installation jobs.