Compute remains on degraded after lock/unlock
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
StarlingX |
Fix Released
|
High
|
zhipeng liu |
Bug Description
Brief Description
-----------------
On Standard Dedicated Storage (2+2+2), compute-1 node remains on degrade status after lock/unlock.
Severity
--------
Provide the severity of the defect.
Major: The other compute remains online.
Steps to Reproduce
------------------
This is part of sanity execution. The actions are in summary:
- system host-lock compute-1
- system host-unlock compute-1
- Monitor progress/status with system host-show compute-1 (node stays on degraded status).
Expected Behavior
------------------
Node compute-1 should be available after lock/unlock operation.
Actual Behavior
----------------
Node stays on degraded status.
Reproducibility
---------------
Seen once. Will update if this appears with newer builds.
System Configuration
-------
Dedicated storage (2+2+2).
Branch/Pull Time/Commit
-------
###
### StarlingX
### Built from master
###
OS="centos"
SW_VERSION="19.01"
BUILD_TARGET="Host Installer"
BUILD_TYPE="Formal"
BUILD_ID=
JOB="STX_
<email address hidden>"
BUILD_NUMBER="207"
BUILD_HOST=
BUILD_DATE=
Last Pass
---------
Build from 2019-08-07.
Timestamp/Logs
--------------
According to the logs, node is on degraded status for the following reason:
| 200. | compute-1 is degraded due to the | host=compute-
| 006 | failure of its 'pci-irq-affinity- | pci-irq-
| | agent' process. Auto recovery of this | | | 006704 |
| | major process is in progress. | | |
Full collect is attached from all nodes.
Test Activity
-------------
Sanity.
Note that this might be related to: https:/
Please re-test with a load built Aug 10 or later as this issue may be fixed via this commit: /review. opendev. org/#/c/ 675503/
https:/