catalog PVC resize with helm

@Madi.Cristil @safiya - please move this to Kasten board for better visibility to the right resources.

@Hagag 

@rajeshvarun Please share the PVC YAML file and the output of the below command,

Please sanitize the output to remove any sensitive information.

 

 helm get values k10 --output yaml --namespace=kasten-io ThanksAhmed Hagag

 helm get values k10 --output yaml --namespace=kasten-io ThanksAhmed Hagag

Hello Hagag, 

  Thanks for your reply.  K10 is deployed via Rancher application. Initially catalog-pvc size is of 100G

pvc.yaml file
Name:          catalog-pv-claim
Namespace:     k10
StorageClass:  default
Status:        Bound
Volume:        pvc-xxxxxx
Finalizers:    [kubernetes.io/pvc-protection]
Capacity:      125Gi     <================Initially it was 100Gi 
Access Modes:  RWO
VolumeMode:    Filesystem
Used By:       catalog-svc-id
Events:        <none>

When I tried to modify the values file and deployed the application via Rancher I got the following error.

 helm get values k10 --output yaml --namespace=k10
auth:
  basicAuth:
    enabled: true
    htpasswd: xxxx
clusterName: xxxx
externalGateway:
  create: true
global:
  cattle:
    clusterId: xxxx
    clusterName: xxxx
    rkePathPrefix: ""
    rkeWindowsPathPrefix: ""
    systemDefaultRegistry: ""
    systemProjectId: xxxxx
    url: https:xxxxx 
  clusterName: xxxxxx
  persistence:
    catalog:
      size: 60Gi
    size: 125Gi                       <=== Increased this from 20Gi to 125Gi in rancher app chart but app is in failed state (not sure why it is 20Gi when actual pvc is 100Gi)
  projectName: xxxx 
  systemDefaultRegistry: ""
ingress:
  annotations:
    cert-manager.io/cluster-issuer: xxxx
  class: xxx
  create: true
  host: xxxxx
  tls:
    enabled: true
    secretName: xxxxx-xxx
injectKanisterSidecar:
  enabled: true
  namespaceSelector:
    matchLabels:
      k10/injectKanisterSidecar: enabled
  objectSelector:
    matchLabels:
      k10/injectKanisterSidecar: enabled
prometheus:
  server:
    persistentVolume:
      enabled: true
secrets:
 xxxx
 xxxxxx
 xxxxxxx

Error: UPGRADE FAILED: cannot patch "k10-grafana" with kind PersistentVolumeClaim: PersistentVolumeClaim "k10-grafana" is invalid: spec.resources.requests.storage: Forbidden: field can not be less than previous value && cannot patch "catalog-pv-claim" with kind PersistentVolumeClaim: PersistentVolumeClaim "catalog-pv-claim" is invalid: spec.resources.requests.storage: Forbidden: field can not be less than previous value

Then I went ahead and increased pvc size directly(100Gi to 125Gi) as am not sure what is causing this issue, space is increased but app in Rancher is still in failed state.

Regards

Raj

@Hagag 

Hello @rajeshvarun,

What was the means of how you increased the Catalog PVC size? Did you change this via attempting to apply a PVC or did you apply this with helm upgrade? 

Thanks

Emmanuel

Hi @EBrockman , Thanks for writing. Earlier I updated the parameter and increased the backend pvc size and deployed the chart resulting in deployment failure but pvc size increased.

“persistence:
        catalog:
          size: 125Gi”

Later I realized that we should be updating following value to update the catalog pvcsize but not above.

“catalog:
              size: 125Gi”

But I still wonder which parameter is pointing to prometheus pvc as I corelated all other volumes expect this.

Regards,

Var

@rajeshvarun Glad that you found out how to customize the helm values in the K10 deployed from rancher app.

Prometheus comes from a different subchart inside k10s chart. So the values used to change the size of the prometheus PVC is a bit different.

It should be prometheus.server.persistentVolume.size

Edit: If the size field is not showing up in your values YAML, You can add it and update it with the size you need.