Giters

topolvm / topolvm

Capacity-aware CSI plugin for Kubernetes

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Topolvm does not provision new logical volume

vasuakkur-s1 opened this issue · comments

commented

Describe the bug
Topolvm version 0.20.0 (https://github.com/topolvm/topolvm/blob/v0.21.0/charts/topolvm/Chart.yaml) was deployed in eks using the legacy settings useLegacy: true (https://github.com/topolvm/topolvm/blob/main/docs/proposals/rename-group.md#user-action) in values files where it retains the provisoner name topolvm.cybozu.com.

We use a node with nvme capbility where the nvme-provisoner creates the volume groups then sets a label on the node nvme=true to topolvm to takeover and create the logicalvolume.When i try to create a new custom application statefulsets the pod gets in pending state, as dynamic provisioning of the pvc does not happen because the logical volume does not exists on the node that pod was assigned.

The logical volumes are not created when setting the helm chart to use the useLegacy: true flag.

Environments

To Reproduce
Steps to reproduce the behavior:

  1. Use the following helm chart https://github.com/topolvm/topolvm/blob/v0.21.0/charts/topolvm/Chart.yaml
  2. updates the values file to use the following
useLegacy: true
controller:
  storageCapacityTracking:
    enabled: true
webhook:
  podMutatingWebhook:
    enabled: false
  1. deploy the above helm chart
  2. the controller deployment and LVM, node daemonsets are running without issues
  3. It does not create a logical volume on the node
    image

image

image
image

Expected behavior
A clear and concise description of what you expected to happen.
The logical volume resource needs to be created.

Additional context
Add any other context about the problem here.

commented

I'm afraid I couldn't reproduce your problem in my local environment. Here's what I tried:

  1. Run git checkout v0.21.0
  2. Edit e2e/manifests/values/daemonset-scheduler-legacy.yaml:
$ cat e2e/manifests/values/daemonset-scheduler-legacy.yaml
scheduler:
  type: daemonset

lvmd:
  managed: false

node:
  lvmdSocket: /tmp/topolvm/lvmd.sock

useLegacy: true
controller:
  storageCapacityTracking:
    enabled: true
webhook:
  podMutatingWebhook:
    enabled: false
  1. Run cd e2e && make start-lvmd && make create-cluster USE_LEGACY=true.
  2. Apply the following manifests to deploy a PVC and a Pod.
$ cat test-pvc1.yaml 
kind: PersistentVolumeClaim
apiVersion: v1
metadata:
  name: test-pvc1
spec:
  accessModes:
  - ReadWriteOnce
  resources:
    requests:
      storage: 1Gi
  storageClassName: topolvm-provisioner

$ cat test-pod1.yaml 
apiVersion: v1
kind: Pod
metadata:
  name: test-pod1
  labels:
    app.kubernetes.io/name: pause
spec:
  containers:
    - name: ubuntu
      image: ubuntu:20.04
      command:
        - bash
        - -c
        - |
          sleep inf &
          trap "kill -SIGTERM $!" SIGTERM
          wait $!
          exit
      volumeMounts:
        - mountPath: /test1
          name: my-volume
  volumes:
    - name: my-volume
      persistentVolumeClaim:
        claimName: test-pvc1

$ bin/kubectl-1.27.3 apply -f test-pvc1.yaml
$ bin/kubectl-1.27.3 apply -f test-pod1.yaml
  1. Check the statuses of the Pod and the PVC.
$ bin/kubectl-1.27.3 get pod
NAME        READY   STATUS    RESTARTS   AGE
test-pod1   1/1     Running   0          95s

$ bin/kubectl-1.27.3 get pvc
NAME        STATUS   VOLUME                                     CAPACITY   ACCESS MODES   STORAGECLASS          AGE
test-pvc1   Bound    pvc-e839127b-9adc-46ac-b264-afbb06632654   1Gi        RWO            topolvm-provisioner   19s

$ bin/kubectl-1.27.3 get pv
NAME                                       CAPACITY   ACCESS MODES   RECLAIM POLICY   STATUS   CLAIM               STORAGECLASS          REASON   AGE
pvc-e839127b-9adc-46ac-b264-afbb06632654   1Gi        RWO            Delete           Bound    default/test-pvc1   topolvm-provisioner            50s

$ sudo lvs
  LV                                   VG          Attr       LSize    Pool Origin Data%  Meta%  Move Log Cpy%Sync Convert
  pool0                                node1-myvg4 twi-a-tz--    4.00g             0.00   11.23                           
  pool0                                node2-myvg4 twi-a-tz--    4.00g             0.00   11.23                           
  d68bd856-9d9c-4635-9021-1eed47189496 node3-myvg1 -wi-ao----    1.00g                                                    
  pool0                                node3-myvg4 twi-a-tz--    4.00g             0.00   11.23                           
  ubuntu-lv                            ubuntu-vg   -wi-ao---- <254.00g

As we can see above, an LV is correctly created.

Could you please explain your problem in more detail? I would appreciate it if you could show me the following:

  • The manifests you used for the Pod and PVC
  • The statuses for the Pod and PVC you deployed (e.g., the output of kubectl describe ...)
commented

This issue has been automatically marked as stale because it has not had any activity for 30 days. It will be closed in a week if no further activity occurs. Thank you for your contributions.

commented

This issue has been automatically closed due to inactivity. Please feel free to reopen this issue (or open a new one) if this still requires investigation. Thank you for your contribution.