Unexpected Deletion of Valid Disks by CloudStack + XCP-ng 8.2.1 #10911
-
We recently encountered an unusual issue with CloudStack, where some valid disks assigned to stopped instances were unexpectedly deleted. Upon investigation, logs from the XCP-ng Center indicated that the delete requests for the VDIs originated from CloudStack. Interestingly, this coincided with the execution of a volume cleanup thread within CloudStack. However, during this cleanup process, CloudStack also issued a delete command for a disk that was actually in a "Ready" state. we could not find any corresponding delete request for that valid disk in the CloudStack management server logs. This raises concerns about a potential issue in the volume cleanup logic or logging mechanisms within CloudStack. Has anyone else experienced a similar scenario, or is aware of what might cause such behavior? audit logs from xen getting requests from cloudstack
Cleanup log reference at the time of incident: |
Beta Was this translation helpful? Give feedback.
Replies: 1 comment 7 replies
-
@midhunpjos Can you search the volume in db by id=2164 ? |
Beta Was this translation helpful? Give feedback.
@midhunpjos
it might be because vm is destroyed on xenserver when ACS stops it.
the behaviour will be changed in 4.19.3 and 4.20.1
#9175