-
Notifications
You must be signed in to change notification settings - Fork 27
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
NumberFormatExceptions followed by "snapshot name is in use for volume ..." #174
Comments
Weird. Since the error seems to be in |
Running with
That last line is interesting... It queries for a backup with Furthermore the snapshot returned is unrecoverable. That should at least unblock me (I can delete it), although I'm surprised it wasn't deleted/cleaned up automatically (it was totally hidden in the CLI, which is as good as deleted, anyway).
What's still confusing me (and where there's maybe room for the CSI to improve) is why this UUID keeps getting recycled for so many different volumes. I know very little about CSI, but naively it seems like each snapshot request should have its own UUID, making collisions like this impossible. The only way I can see that happening is if |
@WanzenBug take a look at LINBIT/linstor-server#314 -- the presence of any backups in the cluster completely breaks snapshotting. |
I'm trying to setup snapshot shipping via CSI (#142) without much success. Most recently I had a flood of errors around parse failures, but there's not enough detail to figure out what it's failing to parse. I thought maybe it had something to do with my VolumeClassName ("linstor.minio") having a period in the name maybe throwing off parsing, but I was able to create another backup (from a significantly smaller volume) using the same class. This all might be a red herring...
It seems after these NumberFormatExceptions happen the system gets into a mixed up state as you can see below -- namely it tries to re-assign an existing snapshot name to a different PVC.
pvc-157c3fce-5720-4ca0-ba11-e2cab638ea93 is indeed another volume that was snapshotted around the same time as this one. Is there perhaps a race condition where creating multiple snapshots around the same time can cause names to not be unique?
# linstor err l
# linstor err show (multiple points)
# linstor err show (For input string)
The text was updated successfully, but these errors were encountered: