action #100853
closed[Timebox: 16h] Investigate iscsi scenarios failing since build 43.1
Description
Observation¶
openQA test in scenario sle-15-SP4-Online-x86_64-autoyast_iscsi_ibft@64bit fails in
installation
According to YaST developers:
(1) we could try the following to identify what is broken:
https://en.opensuse.org/SDB:Linuxrc#p_brokenmodules to check which module breaks it. Or try Manual=2 on Linuxrc commandline.
(2) The logs show a kernel oops.
https://openqa.suse.de/tests/7369836/logfile?filename=serial0.txt
Provide that info to the bug: https://bugzilla.suse.com/show_bug.cgi?id=1191540
Test suite description¶
AutoYaST ibft installation on iSCSI disk. iSCSI configuration is validated in the installed system, as well as cloned profile is verified to check relevant sections.
Reproducible¶
Fails since (at least) Build 43.1
Expected result¶
Last good: 39.1 (or more recent)
Further details¶
Always latest result in this scenario: latest
Updated by JERiveraMoya over 3 years ago
- Project changed from openQA Tests (public) to qe-yam
- Description updated (diff)
- Category deleted (
Bugs in existing tests) - Status changed from New to Workable
Updated by JRivrain over 3 years ago
- Status changed from Workable to In Progress
Updated by JRivrain over 3 years ago
I could not reproduce this manually. I tried updating IPXE kernel, both in my local (outdated) setup and in my local openqa setup, it did the same (failed in openqa, succeeded in virt-manager). According to logs, the kernel oops happens right after detecting IBFT, so my guess would be that this is the faulty module. However, it works in my manual setup, with same IPXE and linux kernels. So that seem to exclude a faulty IBFT module or IPXE kernel.
Next step will be to see what virtual hardware differs between my local VM and openqa's setup, as it seems likely that some module is failing.
Regarding "(1) we could try the following to identify what is broken" there are too many modules loaded at startup, I don't think that's a viable workflow to try excluding them one by one.
I updated https://gitlab.suse.de/qsf-y/qa-sle-functional-y/-/blob/master/Manual_Testing_Guide.md#IBFT with steps to setup a local IBFT environment, in case someone wants to take over.
Note: broken since 43.1, still works in 39.1.
Updated by JERiveraMoya over 3 years ago
- Subject changed from Investigate iscsi scenarios failing since build 43.1 to [Timebox: 16h] Investigate iscsi scenarios failing since build 43.1
- Status changed from In Progress to Blocked
We forgot to timebox the ticket. Kernel developers should care for this, we just needed to investigate a bit.
Let's block it by the bug as we usually do: https://bugzilla.suse.com/show_bug.cgi?id=1191540 and reply there if we are requested for info.
Updated by JRivrain over 3 years ago
In the meantime I could reproduce it manually using the setup described here https://gitlab.suse.de/qsf-y/qa-sle-functional-y/-/blob/master/Manual_Testing_Guide.md#ibft. It did not help me find anything new, but will be useful if we need to provide more info on request.