bitnami / charts

Bitnami Helm Charts
https://bitnami.com
Other
8.85k stars 9.14k forks source link

[bitnami/postgres-ha] persistence.size not being respected #11304

Closed Destreyf closed 2 years ago

Destreyf commented 2 years ago

Name and Version

bitnami/postgres-ha:9.2.4

What steps will reproduce the bug?

Set persistence.size to 20Gi and install the helm chart.

I've tested with a values.yaml and --set and neither work.

Are you using any custom parameters or values?

persistence:
  size: 20Gi

and tried

--set persistence.size=20Gi

What is the expected behavior?

Volume claim should use specified size, in this case it should be 20 Gigabytes.

What do you see instead?

Instance requests an 8Gi PVC which does not work on vultr (they have a 10 Gig minimum)

Name:          data-space-pg-ha-postgresql-ha-postgresql-0
Namespace:     default
StorageClass:  vultr-block-storage
Status:        Pending
Volume:
Labels:        app.kubernetes.io/component=postgresql
               app.kubernetes.io/instance=space-pg-ha
               app.kubernetes.io/name=postgresql-ha
Annotations:   volume.beta.kubernetes.io/storage-provisioner: block.csi.vultr.com
               volume.kubernetes.io/storage-provisioner: block.csi.vultr.com
Finalizers:    [kubernetes.io/pvc-protection]
Capacity:
Access Modes:
VolumeMode:    Filesystem
Used By:       space-pg-ha-postgresql-ha-postgresql-0
Events:
  Type     Reason                Age                   From                                                                             Message
  ----     ------                ----                  ----                                                                             -------
  Normal   ExternalProvisioning  4m47s (x83 over 24m)  persistentvolume-controller                                                      waiting for a volume to be created, either by external provisioner "block.csi.vultr.com" or manually created by system administrator
  Warning  ProvisioningFailed    74s (x14 over 24m)    block.csi.vultr.com_csi-vultr-controller-0_d3e281b6-a9f7-4eeb-a469-b08807cd4480  failed to provision volume with StorageClass "vultr-block-storage": rpc error: code = Internal desc = {"error":"Unable to add block storage: Failed to add block storage device. Invalid size (8GB) - Must be between 10GB and 10000GB.","status":400}
  Normal   Provisioning          7s (x15 over 24m)     block.csi.vultr.com_csi-vultr-controller-0_d3e281b6-a9f7-4eeb-a469-b08807cd4480  External provisioner is provisioning volume for claim "default/data-space-pg-ha-postgresql-ha-postgresql-0"

Additional information

I am using Vultr's VKE environment and they have a minimum nvme block size of 10 gigabytes, and a minimum hdd block size of 40 gigabytes.

Destreyf commented 2 years ago

Ignore this, I didn't realize that when i tore down the services the volumes remained, so my original deployment attempt had left existing volumes there, once I removed the original failed volumes they were re-provisioned correctly.