Project

General

Profile

coordination #105082

coordination #105073: [Epic] Improve logging in openQA

Evaluate more simple and consistent ways to collect logs

Added by JRivrain 8 months ago. Updated 7 months ago.

Status:
New
Priority:
Low
Assignee:
-
Target version:
Start date:
2022-01-19
Due date:
% Done:

0%

Estimated time:
Tags:

Description

To be done after Identify which post_fail_hook is used by each module in Yast group, and what it does.

Currently the logging in openQA is a often chaotic, sometimes redundant, sometimes some things are missing. Also, some of the files collected contain no information at all. We should try to make logging more consistent.

We would like to evaluate different ways of collecting logs.

  • supportconfig: This tool does most of what we are doing in post_fail_hooks (and more) but more consistently. Some devs may be more familiar with it.
  • journalctl: journalctl is now the default for almost all linux distributions. Yet, for most cases, we don't seem to have the output of journalctl in logs.
  • tar: a simple tarball of /var/log is simpler than collecting files individually, and we are sure not to miss any log file.

Proposals:

  • read manpage of supportconfig. some options can be useful, like -l (all lines), and the possiblitity to upload directly the output. However, for Yast, it does not seem to contain the rotated logs (eg y2log2.gz).
  • for journalctl, we can just use journalctl > journalctl.log, then upload it.

AC1: Identify useless, redundant, or in the contrary, missing log collections for each module in Yast group.
AC2: Make proposal(s) on how to replace current log collection by something more consistent.


Related issues

Related to qe-yast - action #105079: [Research][timebox: 24] Evaluate a way to snapshot and upload the entire system at the point of failure, that could be bootable as VM.Rejected2022-01-19

History

#1 Updated by JRivrain 8 months ago

  • Description updated (diff)

#2 Updated by JRivrain 8 months ago

  • Description updated (diff)

#3 Updated by JRivrain 8 months ago

  • Description updated (diff)

#4 Updated by JRivrain 8 months ago

  • Related to action #105079: [Research][timebox: 24] Evaluate a way to snapshot and upload the entire system at the point of failure, that could be bootable as VM. added

#5 Updated by JRivrain 8 months ago

  • Tags set to qe-yast-refinement

#6 Updated by JERiveraMoya 8 months ago

supportconfig seems to be a tool for the installed system: https://documentation.suse.com/sles/15-SP3/html/SLES-all/cha-adm-support.html
journalctl is not retrieved always, but it is there many time, maybe it is a more general issue about retrieving logs, perhaps we need some examples.
tar the whole /var/log might not be realiable, any process can be written there, i.e.: warnings of the kind "File changed as we read". Also we should take into account the size.

#7 Updated by JERiveraMoya 8 months ago

  • Target version set to Current

#8 Updated by JERiveraMoya 7 months ago

  • Priority changed from Normal to Low

Also available in: Atom PDF