Preparing Disks for LINSTOR Storage Pools

This guide explains how to prepare physical disks for use with LINSTOR when they contain old metadata that prevents automatic detection.

Problem Description

When setting up storage on new or repurposed nodes, physical disks may contain remnants from previous installations:

• RAID superblocks
• Partition tables
• LVM signatures
• Filesystem metadata

This old metadata prevents LINSTOR from detecting disks as available storage.

Symptoms

1. linstor physical-storage list shows empty output or missing disks
2. Disks appear with unexpected filesystem types (e.g., linux_raid_member)
3. Storage pools only show DfltDisklessStorPool without actual storage

Diagnostics

Set up LINSTOR alias

For easier access to LINSTOR commands, set up an alias:

alias linstor='kubectl exec -n cozy-linstor deploy/linstor-controller -- linstor'

Check LINSTOR nodes

List your nodes and check their readiness:

linstor node list

Expected output should show all nodes in Online state.

Check storage pools

Check current storage pools:

linstor storage-pool list

Check available physical storage

Check what physical disks LINSTOR can see:

linstor physical-storage list

If this command shows empty output or is missing expected disks, the disks likely contain old metadata and need to be wiped.

Check disk state on node

Check disk state on a specific node via satellite pod:

# List LINSTOR satellite pods
kubectl get pod -n cozy-linstor -l app.kubernetes.io/component=linstor-satellite

# Check disk state
kubectl exec -n cozy-linstor <satellite-pod-name> -c linstor-satellite -- \
  lsblk -f

Expected output for clean disks should show no FSTYPE:

NAME    FSTYPE LABEL UUID MOUNTPOINT
nvme0n1
nvme1n1

If you see linux_raid_member, LVM2_member, or other filesystem types, the disks need to be wiped.

Solution: Wiping Disk Metadata

Step 1: Identify System Disks

Before wiping, identify which disk contains your Talos installation. Check your Talos configuration in nodes/<node-name>.yaml:

# In nodes/<node-name>.yaml
machine:
  install:
    disk: /dev/sda  # This disk should NOT be wiped

Typically, the system disk is /dev/sda, /dev/vda, or similar.

Step 2: Locate Node Configuration

If you used Talm to bootstrap your cluster, your node configurations are stored in nodes/*.yaml files in your cluster configuration directory.

Each file corresponds to a specific node (e.g., nodes/node1.yaml, nodes/node2.yaml).

Step 3: Wipe Disks

List all disks on the node:

talm -f nodes/<node-name>.yaml get disks

Wipe all non-system disks:

talm -f nodes/<node-name>.yaml wipe disk nvme0n1 nvme1n1 nvme2n1 ...

Step 4: Verify Disks are Clean

After wiping, verify that disks are now visible to LINSTOR:

linstor physical-storage list

Expected output should now show your disks:

+----------------------------------------------------------------+
| Device    | Size       | Rotational |
|================================================================|
| /dev/nvme0n1 | 3.49 TiB   | False      |
| /dev/nvme1n1 | 3.49 TiB   | False      |
| ...          | ...        | ...        |
+----------------------------------------------------------------+

You can also check directly on the node:

kubectl exec -n cozy-linstor <satellite-pod-name> -c linstor-satellite -- \
  lsblk -f

Clean disks should show no FSTYPE.

Creating Storage Pools

Once disks are clean, create a LINSTOR storage pool.

For ZFS storage pools with multiple disks:

linstor physical-storage create-device-pool zfs <node-name> \
  /dev/nvme0n1 /dev/nvme1n1 /dev/nvme2n1 ... \
  --pool-name data \
  --storage-pool data

Verify the storage pool was created:

linstor storage-pool list

Expected output:

+-----------------------------------------------------------------------+
| StoragePool | Node  | Driver | PoolName | FreeCapacity | TotalCapacity | State |
|=======================================================================|
| data        | node1 | ZFS    | data     | 47.34 TiB    | 47.62 TiB     | Ok    |
| data        | node2 | ZFS    | data     | 47.34 TiB    | 47.62 TiB     | Ok    |
+-----------------------------------------------------------------------+

Troubleshooting

Disks Still Show Old Metadata After Wipe

Try wiping with the ZEROES method for more thorough cleaning:

talm -f nodes/<node-name>.yaml wipe disk --method ZEROES nvme0n1

This writes zeros to the disk, which takes longer but ensures complete removal of metadata.

“Zpool name already used” Error

If you need to recreate a storage pool:

1. Delete from LINSTOR:

   linstor storage-pool delete <node-name> <pool-name>
   

2. Destroy ZFS pool on the node:

   kubectl exec -n cozy-linstor <satellite-pod-name> -c linstor-satellite -- \
     zpool destroy <pool-name>
   

3. Recreate the pool with all disks in one command.

Permission Denied on Worker Nodes

Worker nodes may not allow direct Talos API access. Use the satellite pod to check disk state:

kubectl exec -n cozy-linstor <satellite-pod-name> -c linstor-satellite -- lsblk -f

If you need to wipe disks on worker nodes, ensure your node configuration allows access or consult your cluster administrator.

Quick Reference

Related Documentation

• Using Talm to Bootstrap Cozystack
• Configuring a Dedicated Network for LINSTOR
• Configuring DRBD Resync Controller
• LINSTOR Troubleshooting