b3d86f4033
Now that we have better logging in tempest via commit d2b9651cb82910b2b86b0f94af1721f717e74024, we can update the query to include the volume delete timeout message from the tempest logs. This adds another 27 hits in the last 24 hours to the failure count, which should get our gate categorization percentage up from the measly 58% we have now. As for the actual signatures, we're basically treating bug 1373513 as the home for all of the current cinder blocking lvm/lvs/vgs issues right now. When we see the volume delete timeout message in tempest we don't always see the lvm message in syslog, so it could be problems with vgs taking too long also (we know that's an intermittent problem), but we suspect the root issues are the same in Cinder and need to be investigated together. Change-Id: Ia100f7c5174a9629a90df7b13030b88195baac12 Related-Bug: #1373513
8 lines
217 B
YAML
8 lines
217 B
YAML
query: >-
|
|
(message:"INFO: task lvm:" AND
|
|
message:"blocked for more than 120 seconds" AND
|
|
tags:"syslog") OR
|
|
(message:"Failed to delete volume" AND
|
|
message:"within the required time" AND
|
|
tags:"console")
|