PersistentVolume documentation contradicts itself

Couchbase Server Kubernetes
1

If you look at the docs for the helm chart persistent volumes , the example includes a snippet about defining volume mounts via

  servers:
    default:
      pod:
        volumeMounts:
          default: couchbase
          data:  couchbase

and then a link in the additional values below this links to a schema defined as…

  servers:
  - size: 1
    name: all_services
    volumeMounts:
      default: couchbase
      data:  couchbase
      index: couchbase
      analytics:
      - couchbase
      - couchbase

so which is it?

2

It’s the 1st example there. The helm template is not necessarily formatted the same as the couchbasecluster crd. Helm uses a map representation of the server config and then internally converts it to array representation of the 2nd example. The reason for this is overrides are difficult to do with the array format… ie:
--set cluster.servers[1].size=3
Whereas map is more explicit approach of --set cluster.servers.default.size=3

3

Also, the operator values.yaml file shows that volumeMounts is a sibling to pod unlike the first example above. So far it seems that none of these are working with the volumeClaimTemplates