Steps to restore backed-up data using Longhorn, with a PostgreSQL example.

longhorn_restore_guide.md

Restore Backup with Longhorn

Instructions for restoring data using Longhorn backup and restore features.

Restore Backup

Example: Restoring PostgreSQL Data

1. Scale Down the PostgreSQL StatefulSet:

   kubectl scale statefulset.apps/postgresql --replicas=0 -n postgresql

2. Wait for the Volume to Detach:

   • Monitor the volume status until it is detached.

3. Note the Current Volume Name:

   • For example: pvc-1be4dafb-399f-40d4-ac87-205eb56c2f44.

4. Delete the Old Volume via Longhorn GUI:

   • Navigate to the Longhorn GUI and delete the specified volume.

5. Restore the Backup to a New Volume:

   • Use the Longhorn GUI to restore the backup to a new volume.
   • Assign the new volume the same name noted in step 3 (pvc-1be4dafb-399f-40d4-ac87-205eb56c2f44).

6. Wait for the Restore Process to Complete:

   • Monitor the Longhorn GUI until the restore process finishes.

7. Create Persistent Volume (PV) / Persistent Volume Claim (PVC):

   • Navigate to the lonhorn GUI > Volume > at the right side of volume named pvc-1be4dafb-399f-40d4-ac87-205eb56c2f44 open operation button > select Create PV/PVC
   • Ensure the Create PVC option is checked.
   • Ensure the Use Previous PVC option is checked.

8. Wait for the PV/PVC to be Available:

   • Monitor the status until the PV/PVC is available and ready for use.

9. Scale Up the PostgreSQL StatefulSet:

   kubectl scale statefulset.apps/postgresql --replicas=1 -n postgresql

By following these steps, you can successfully restore your data using Longhorn.

How to scale down and up a postgresql statefulset :

kubectl scale statefulset.apps/postgresql --replicas=0 -n postgresql
kubectl scale statefulset.apps/postgresql --replicas=1 -n postgresql

How disable auto-sync in the argocd application and then scale it down and up :

1- Select the application and click on the DETAILS button up left corner :

2- At the Sync Policy section, hit the DISABLE AUTO-SYNC :

3- Now you can easily manage to scale down or up :

kubectl scale deployment.apps/storage --replicas=0 -n app-production-storage
kubectl scale deployment.apps/storage --replicas=1 -n app-production-storage

How to force a pvc to be deleted :

1- Read the pvc's detail :

kubectl get pvc storage-pv-claim -n app-production-storage -o yaml

2- Set it's finalizers to null :

kubectl patch pvc storage-pv-claim -p '{"metadata":{"finalizers":null}}' -n app-production-storage

3- Now you can delete it easily :

kubectl delete pvc storage-pv-claim -n app-production-storage