Add secondary disks to Longhorn programatically/without the UI #11909

JimFawkes · 2025-10-03T11:39:49Z

JimFawkes
Oct 3, 2025

Hello,
I have several Nodes in my cluster with secondary storage disks attached. I would now like to make the available to longhorn. I am able to do this from the UI (see screenshot below), however I would like to avoid manual changes in UIs and define the disk setup in Code. Unfortunately I have not been able to do this so far. What are the recommended ways of achieving this?

I have attempted to patch the Nodes with a kustomize patch as follows:

  - patch: |-
      - op: add
        path: /spec/disks/storage
        value:
          allowScheduling: true
          diskType: filesystem
          path: /var/mnt/storage
    target:
      version: longhorn.io/v1beta2
      kind: Node
      name: mynode-n2
      namespace: longhorn-system

However this is not applied.

Any help or links to resources that could help are very much appreciated.

Disk successfully added in the UI:

mantissahz · 2025-10-07T01:36:58Z

mantissahz
Oct 7, 2025
Collaborator

define the disk setup in Code.

Which language do you want to use?

I have attempted to patch the Nodes with a kustomize patch as follows:

  - patch: |-
      - op: add
        path: /spec/disks/storage
        value:
          allowScheduling: true
          diskType: filesystem
          path: /var/mnt/storage
    target:
      version: longhorn.io/v1beta2
      kind: Node
      name: mynode-n2
      namespace: longhorn-system

Could you provide the Longhorn node information after you path the node with a kustomize patch?

kubectl -n longhorn-system get lhn mynode-n2 -oyaml

0 replies

JimFawkes · 2025-10-15T09:11:18Z

JimFawkes
Oct 15, 2025
Author

Hi,

define the disk setup in Code.

Which language do you want to use?

Hmm probably not the best choice of words on my side. I just want to define it in any way that I can push to version control. So any yaml manifest helm values api calls to make in a script or terraform resources to define would be ok. ATM I am using ArgoCD with kustomize and helm to deploy applications and services to my cluster.

Could you provide the Longhorn node information after you path the node with a kustomize patch?
kubectl -n longhorn-system get lhn mynode-n2 -oyaml

Sure, here is the output:

apiVersion: longhorn.io/v1beta2
kind: Node
metadata:
  creationTimestamp: "2025-10-02T19:14:09Z"
  finalizers:
    - longhorn.io
  generation: 2
  name: mynode-n2
  namespace: longhorn-system
  resourceVersion: "4744423"
  uid: 053...afb
spec:
  allowScheduling: true
  disks:
    default-disk-b30400000000:
      allowScheduling: true
      diskDriver: ""
      diskType: filesystem
      evictionRequested: false
      path: /var/lib/longhorn/
      storageReserved: 37960758067
      tags: []
    storage:
      allowScheduling: true
      diskDriver: ""
      diskType: filesystem
      evictionRequested: false
      path: /var/mnt/storage
      storageReserved: 0
      tags: []
  evictionRequested: false
  instanceManagerCPURequest: 0
  name: mynode-n2
  tags: []
status:
  autoEvicting: false
  conditions:
    - lastProbeTime: ""
      lastTransitionTime: "2025-10-05T14:33:37Z"
      message: Node mynode-n2 is ready
      reason: ""
      status: "True"
      type: Ready
    - lastProbeTime: ""
      lastTransitionTime: "2025-10-02T19:14:09Z"
      message: ""
      reason: ""
      status: "True"
      type: Schedulable
    - lastProbeTime: ""
      lastTransitionTime: "2025-10-02T19:14:09Z"
      message: All required packages [dmsetup nfs-common iscsiadm cryptsetup] are installed
      reason: ""
      status: "True"
      type: RequiredPackages
    - lastProbeTime: ""
      lastTransitionTime: "2025-10-02T19:14:09Z"
      message: ""
      reason: ""
      status: "True"
      type: Multipathd
    - lastProbeTime: ""
      lastTransitionTime: "2025-10-02T19:14:09Z"
      message: ""
      reason: ""
      status: "True"
      type: NFSClientInstalled
    - lastProbeTime: ""
      lastTransitionTime: "2025-10-02T19:14:09Z"
      message: Kernel modules [dm_crypt] are loaded
      reason: ""
      status: "True"
      type: KernelModulesLoaded
    - lastProbeTime: ""
      lastTransitionTime: "2025-10-02T19:14:39Z"
      message: ""
      reason: ""
      status: "True"
      type: MountPropagation
  diskStatus:
    default-disk-b30400000000:
      conditions:
        - lastProbeTime: ""
          lastTransitionTime: "2025-10-02T19:14:39Z"
          message:
            Disk default-disk-b30400000000(/var/lib/longhorn/) on node mynode-n2
            is ready
          reason: ""
          status: "True"
          type: Ready
        - lastProbeTime: ""
          lastTransitionTime: "2025-10-02T19:14:39Z"
          message:
            Disk default-disk-b30400000000(/var/lib/longhorn/) on node mynode-n2
            is schedulable
          reason: ""
          status: "True"
          type: Schedulable
      diskDriver: ""
      diskName: default-disk-b30400000000
      diskPath: /var/lib/longhorn/
      diskType: filesystem
      diskUUID: bf3...51a
      filesystemType: xfs
      instanceManagerName: instance-manager-5d2...662
      scheduledBackingImage: {}
      scheduledReplica: {}
      storageAvailable: 116706508800
      storageMaximum: 126535860224
      storageScheduled: 0
    storage:
      conditions:
        - lastProbeTime: ""
          lastTransitionTime: "2025-10-03T11:36:09Z"
          message: Disk storage(/var/mnt/storage) on node mynode-n2 is ready
          reason: ""
          status: "True"
          type: Ready
        - lastProbeTime: ""
          lastTransitionTime: "2025-10-09T15:18:24Z"
          message: Disk storage(/var/mnt/storage) on node mynode-n2 is schedulable
          reason: ""
          status: "True"
          type: Schedulable
      diskDriver: ""
      diskName: storage
      diskPath: /var/mnt/storage
      diskType: filesystem
      diskUUID: 916...11a
      filesystemType: xfs
      instanceManagerName: instance-manager-5d2...662
      scheduledBackingImage: {}
      scheduledReplica:
        pvc-6e4...53b-r-b1b...4a7: 53687091200
        pvc-6ef...f25-r-fbc...e5e: 2147483648
        pvc-49c...4b3-r-770...062: 53687091200
        pvc-b5a...da2-r-73f...a9b: 53687091200
        pvc-d49...654-r-ac4...2cb: 26843545600
      storageAvailable: 497129881600
      storageMaximum: 511858737152
      storageScheduled: 190052302848
  region: ""
  snapshotCheckStatus:
    lastPeriodicCheckedAt: "2025-10-15T00:00:00Z"
  zone: ""

0 replies

rgarrigue · 2026-02-20T21:25:07Z

rgarrigue
Feb 20, 2026

I've the exact same question. I wish there would be something in the Helm chart, a dedicated field or even a longhorn resource manifest that I would add in extraObjects. Or a label on the node, whatever.

0 replies

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Add secondary disks to Longhorn programatically/without the UI #11909

Uh oh!

{{title}}

Uh oh!

Replies: 3 comments

Uh oh!

{{title}}

Uh oh!

Uh oh!

{{title}}

Uh oh!

Uh oh!

{{title}}

Uh oh!

Select a reply

Uh oh!

Add secondary disks to Longhorn programatically/without the UI #11909

Uh oh!

JimFawkes Oct 3, 2025

Replies: 3 comments

Uh oh!

mantissahz Oct 7, 2025 Collaborator

Uh oh!

JimFawkes Oct 15, 2025 Author

Uh oh!

rgarrigue Feb 20, 2026

JimFawkes
Oct 3, 2025

mantissahz
Oct 7, 2025
Collaborator

JimFawkes
Oct 15, 2025
Author

rgarrigue
Feb 20, 2026