Randomly checkpoint inconsistent: missing or broken bitmap #267
Replies: 6 comments
-
|
This can happen if vm is powered off without qemu beeing able to write the bitmaps or if vm is migratet to another host, check readme in this case, or your checkpoint directory isnt in sync with the checkpoints |
Beta Was this translation helpful? Give feedback.
-
|
Or some other component Alters the vm disks images (resize)? In between backups |
Beta Was this translation helpful? Give feedback.
-
|
Hi, I read the readme. And none of the things mentioned therein happened. Disk didn't resize either. |
Beta Was this translation helpful? Give feedback.
-
|
Use qemu-img to check bitmap state.. no complains by other users so I guess it’s a issue caused by other components |
Beta Was this translation helpful? Give feedback.
-
|
maybe opennebula re-deploys (undefine/define) virtual machine each time it is powered off and as such checkpoint/bitmap information is lost. |
Beta Was this translation helpful? Give feedback.
-
|
Happens to me and it correlates to sudden turn off of the VM (sudden freeze of the server). So I understand the explanation.
This way I have a weekly rotation. Which is great to keep NVME wear to minimum, but not great to have as many backups as possible, it's only up to 7 (from 1 to 7). I have no space to do what "auto" mode intents to do in the documentation - make a new folder and continue snapshots. |
Beta Was this translation helpful? Give feedback.
Uh oh!
There was an error while loading. Please reload this page.
-
Version used
2.26
Describe the bug
I use virtnbdbackup at several nodes, and many VM are backed up correctly, but sometimes I get the error:
Redefining checkpoint failed: [virtnbdbackup.0]: checkpoint inconsistent: missing or broken bitmap 'virtnbdbackup.0' for disk 'vda'That happens randomly in about 5-10% of VMs. I can't find any common features between the problematic VMs. I checked if the VMs were migrated or unexpected shutdown, but not.
Command what I run:
/usr/local/bin/virtnbdbackup -d one-2080 -o /srv/backup/2025-W15/one-2080 -l auto --noprogress --checkpointdir /srv/backup/checkpoints/one-2080 --verboseContent of backups and checkpoints directory:
But there are no checkpoints:
Expected behavior
Could you provide some troubleshooting steps to help resolve the problem?
Hypervisor and/or virtual machine information:
Logfiles:
Workaround:
After removing backups and checkpoints dirs backups become possible.
Beta Was this translation helpful? Give feedback.
All reactions