action #12064
closedImproved logging for debugging performance related issues
0%
Description
User story¶
As a developer trying to fix sporadic or potential performance issues, I want access to more logs and data from the SUT to be able to distinguish "openQA worker" performance impact from performance regressions in the product
acceptance criteria¶
- Logfiles with blocked tasks and system load is available in case of sporadic failures
tasks¶
- Call "uptime" in case of error which gives information about load, too, e.g.
13:57 up 16 days 1:09, 0 users, load average: 0.25, 0.42, 0.48or callcat /proc/loadavg - Optional: Repeatedly call the above over some time, e.g. before and after every test
further details¶
related to #11928
Updated by okurz over 8 years ago
- Related to action #12110: gnome_control_center: short timeout on details caused test to fail added
Updated by okurz over 8 years ago
PR ready: https://github.com/os-autoinst/os-autoinst-distri-opensuse/pull/1393
the "psaxf" should be enough to see if any processes are blocked or is more necessary?
Updated by okurz over 8 years ago
- Status changed from In Progress to Feedback
- Assignee deleted (
okurz)
PR is merged, let's see if anything else comes up.
Updated by okurz almost 8 years ago
- Status changed from Feedback to Resolved
- Assignee set to okurz
we improved a lot in this regard since the last comment. coolo added performance checks for check/assert_screen call durations and also some others. Tests abort when there is a timeout on check/assert_screen and previously delays have been encountered. Also, IO related logs are available in the logfiles.
Updated by okurz almost 8 years ago
- Related to coordination #14972: [tools][epic] Improvements on backend to improve better handling of stalls added