Ability to add additional volumes to Zookeeper Pods

[hmontesdeoca]

Hello,

I am currently using Azure to set up aks clusters utilizing solr-cloud helm charts and the provided zookeeper ensemble. I want to mount zookeeper logs onto a Azure File Share but I cant seem to specify a pvc for the zk provided ensemble. It only seems to accept a 'spec' as some sort of reference. How can I specify a PVC for zookeeper? Can I mount directly to the logs directory for zookeeper pods?

 zk:
    provided:
      persistence:
        spec:
          storageClassName:  "solr"
          #can I specify a path, volume, volume mount, or pvc?

[Frankkkkk]

You must specify a storage class name. You can list them with kubectl get sc. On azure, you could use azurefile-csi-premium for example

[hmontesdeoca]

I have specified a storage class name but I would like to mount this pvc to an azure file share. For example, once I create this release via helm I notice it automatically creates new data pvc's according to how many replicas I would like set for the ensemble.

In this case the following command kubectl get pvc -n <namespace> with 3 replicas for a values file this would show:

<release-name>-solrcloud-zookeeper-0
<release-name>-solrcloud-zookeeper-1
<release-name>-solrcloud-zookeeper-2

And they are all bound to the storage class "solr" as mentioned above with the snippet of the yaml file.
I would simply like to have a mount to a fileshare represented as a pvc in addition to these automatically provisioned pvcs. This could be seen with the solrcloud helm chart and many others. I just cant seem to find the documentation on this zookeeper provided subchart when specifying volume mounts.

[HoustonPutman]

Ahh so it looks like the zookeeper operator does support additional volumes, but the solr operator doesn't allow that functionality. We should make a PR to add the additional functionality not supported by the Solr Operator.

[hmontesdeoca]

thank you for the information, I will relay this to my team.

[HoustonPutman]

This is going to be handled in #557