Repository navigation
Error while trying to create a volume from a volume snapshot that sometimes result in snapshot loss #9538
Closed as not planned
Description
ISSUE TYPE
- Bug Report
COMPONENT NAME
KVM, Volume Snapshot
CLOUDSTACK VERSION
4.19.1.0
CONFIGURATION
OS / ENVIRONMENT
1 Zone, 1 Pod, 1 Cluster
Advanced networking
1 KVM host (Ubuntu 24.04)
1 Management Server (Ubuntu 24.04)
NFS primary storage
SUMMARY
Trying to create a volume from a volume snapshot results in an error that sometimes removes the snapshot.
STEPS TO REPRODUCE
When snapshot.backup.to.secondary is disabled, the following procedure sometimes results in an error at the last step:
- Take volume snapshot A of VM X
- Take volume snapshot B of VM X
- Create a volume from snapshot A
- Create a volume from snapshot B
When this happens, the second volume is not created, and sometimes volume snapshot B is removed from the primary storage.
EXPECTED RESULTS
Volume is created from volume snapshot
ACTUAL RESULTS
Volume is not created, and the snapshot is sometimes lost
Activity
[-]Error while trying to create a volume from the snapshot that sometimes result in snapshot loss[/-][+]Error while trying to create a volume from a volume snapshot that sometimes result in snapshot loss[/+]DaanHoogland commentedon Nov 8, 2024
@winterhazel as #9270 is still being discussed, I am marking this for 4.21 for now. cc @slavkap @JoaoJandre
nvazquez commentedon Jul 11, 2025
Hi @winterhazel - is this issue still applicable now that #9270 is merged? If not, can it be closed?
winterhazel commentedon Jul 12, 2025
@nvazquez yes, it is still applicable. #9270 does not handle this issue.
19 remaining items
DaanHoogland commentedon Mar 5, 2026
@winterhazel , is this relevant for 22 onwards? (can we still reproduce?) cc @sureshanaparti .
github-actions commentedon Aug 31, 2026
This issue is stale because it has been open for 120 days with no activity. It may be removed by administrators of this project at any time. Remove the stale label or comment to request for removal of it to prevent this.
DaanHoogland commentedon Aug 31, 2026
ping @winterhazel @sureshanaparti @abh1sar @vladimirpetrov @harikrishna-patnala Is this serious enough to persue in the coming months?
winterhazel commentedon Sep 10, 2026
@DaanHoogland I think we can close this one. I am not able to reproduce the snapshot loss anymore, and the copy failure seems more like a permissioning issue with the environment I was using at that time.