# Longhorn — Backup & Restore Volumes

# 1\. Prerequisites

Longhorn installed and healthy (longhorn-system pods ready).

A Backup Target reachable from all Longhorn manager pods:

S3-compatible bucket (with access/secret and optional custom endpoint), or

NFS export (read/write, stable network).

Sufficient free space on the BackupStore.

![](https://cdn.hashnode.com/res/hashnode/image/upload/v1762245379727/e46a9708-1685-4659-875e-344aefb384d5.png align="center")

# 2.Set the Backup Target

UI (recommended)

1. Open Longhorn UI → Settings → General.
    
2. Set Backup Target (examples):
    

* S3: s3://my-bucket@us-east-1/longhorn
    
* NFS: nfs://10.0.0.20:/export/longhorn-backups
    

3. If using S3, set Backup Target Credential Secret (Access Key, Secret, Region, Endpoint if non-AWS).
    

kubectl (advanced)

* Create a Secret with S3 creds in longhorn-system, then patch Settings (keys vary by version). In most cases the UI is simpler and safer.
    

# 3\. Create Snapshots & Backups

Manual (per volume)

1. Longhorn UI → Volumes → click your volume.
    
2. Snapshots tab → Create Snapshot (name it).
    
3. Backups tab → Create Backup from the snapshot (or directly “Backup”).
    

Recurring (recommended)

* Per volume → Recurring Jobs:
    
* * Hourly/Daily Snapshot (e.g., keep last 7)
        
* * Daily/Weekly Backup (e.g., keep last 30)
        
* Attach the recurring job(s) to the volume (or set globally and opt-in).
    

# 4\. Verify Backups

UI

* Volume → Backups tab → you should see backup entries with timestamps and sizes.
    
* Click a backup to view details (labels, CRC, etc.).
    

Best practice

* Periodically restore a backup to a scratch/test volume and run a quick read/write check.
    

# 5\. Restore a Backup (New Volume)

UI

1. Longhorn UI → Backups (left menu).
    
2. Choose the Backup Volume (it groups backups by original volume).
    
3. Pick the desired Backup → Restore.
    
4. Set New Volume Name, size (auto from backup), Data Engine, # of replicas, and (optional) Create PV/PVC.
    
5. Click OK. Longhorn creates the volume and starts rebuilding replicas.
    
6. When the volume is Healthy, Attach it to a node, then mount it via your workload (or use the auto-created PVC).
    

Notes

* If you created PV/PVC automatically, just point your Pod/Deployment/StatefulSet to that PVC.
    
* File system (ext4/xfs) will match the original backup unless you override.
    

# 6\. Disaster Recovery (Cross-cluster or Failover)

DR Volume approach

1. On the target cluster, configure the same Backup Target.
    
2. Longhorn UI → Backups → select the source Backup Volume.
    
3. Click Create Disaster Recovery Volume (DR Volume).
    
4. Longhorn creates a DR volume that periodically pulls incremental backups.
    
5. If the source cluster goes down, Activate the DR volume:
    

* DR Volume → Activate → it becomes a regular volume.
    
* Attach, and use like normal.
    

When to use

* For warm standby and faster RTO than restoring from scratch.