-
Notifications
You must be signed in to change notification settings - Fork 7
Unable to restore from backups #67
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
Comments
There's errors in there, but none that are particularly clear about what's going on (as far as I can see). Can you give some more detail about what the Cluster resource does, jobs/pods that are going around, logs from the cnpg operator, etc? |
I think this is the problem of the name being too long? |
We have patched it in tensorchord/pgvecto.rs#588. However updating version with immich database is quite complicated. We'll work with immich when we have a new production release ready |
Encountered this issue today after deleting a cloudpg pod & pvc.
Thanks for for patching. It looks like the fix is in 0.4.0 but as yet immich only supports < 0.4.0. I tried, it wouldn't start 😢 . I have database dumps for a full restore so all is good. Just wondering if it would it be possible to fix on 0.3.0 also? Thanks! |
@beyondwatts Are you experiencing problem at restore or backup? |
This was after deleting a pod & pvc in the cluster so I guess restoring |
Came here after running into the same issue, for the same reason (immich only supporting <0.4.0) and my HA postgres deployment can't scale or recover a secondary node because of this error. |
Seeing the same issue, even tried recovering with 0.4.0 and same error. I suspect because that fix is for going forward? Is the only way forward to recover to a single pod instance (which works fine) Create a new cluster and do pg_dumps from the old to the new one? |
@hagak You can just drop all the vector index and rebuild them. |
Any info on this procedure? Do I just drop the index via sql on the single instance that came up and the replicas will then start up? After that just add the vector index back? |
@VoVAllen So I deployed a single CNPG instance from backup and it starts. I then deleted the 2 vector indexes I could find in immich db. Then tried to deploy a replicate instance and same file name too long error. The two indexes i found are under the "face_search" and "smart_search" tables. Did I miss one? |
@hagak, I'm not a db expert and used the help of grok to resolve this issue. Make sure you have backups. For context, this is the error i was getting when trying to rebuild my cluster:
i started my cluster with 1 instance, let it rebuild from s3, then i searched for indexes with:
and got the following output:
I dropped the 2 indexes with:
i was then able to scale up the number of instances without error. After verifying immich was running as expected, i recreated the indexes with:
and
If anyone here, with more knowledge, notices anything wrong about this, please let me know so I can correct it. |
Unfortunately it seems that the backed up archives are not able to be restored..
ghcr.io/tensorchord/cloudnative-pgvecto.rs:16.2-v0.2.1
Cluster CRD:
The text was updated successfully, but these errors were encountered: