Disable persistent does not work

[soroshsabz]

ITNOA

I try to create redis enterprise cluster with below declaration

apiVersion: "app.redislabs.com/v1"
kind: "RedisEnterpriseCluster"
metadata:
  name: "harbor-cluster"
spec:
  nodes: 3
  persistentSpec:
    enabled: false
  redisEnterpriseNodeResources:
    limits:
      cpu: 1000m
      memory: 1Gi
    requests:
      cpu: 1000m
      memory: 1Gi

But my problem is even I set presistentSpec to false, I see kubectl get pvc show redis try to claim pv and my bootstrapping of my pods is failed.

my question is how to resolve it?

[soroshsabz]

related to #177

[laurentdroin]

Hi,

That seems very strange. Would you be able to run the log_collector.py script in order to generate a diagnostic package and open a Support Ticket with Redis so that we can take a look?
Also, the memory requested (1Gi) is too small. The nodes will not work. We recommend at least 4Gi for testing. I think 3 Gi "might" work.

Laurent.

[soroshsabz]

My problem is why

kubectl get pvc has result for redis?

If I run kubectl describe pvc redis-enterprise-storage-harbor-cluster-0 we can see below result

Name:          redis-enterprise-storage-harbor-cluster-0
Namespace:     default
StorageClass:
Status:        Pending
Volume:
Labels:        app=redis-enterprise
               redis.io/cluster=harbor-cluster
               redis.io/role=node
Annotations:   <none>
Finalizers:    [kubernetes.io/pvc-protection]
Capacity:
Access Modes:
VolumeMode:    Filesystem
Used By:       harbor-cluster-0
Events:
  Type    Reason         Age                      From                         Message
  ----    ------         ----                     ----                         -------
  Normal  FailedBinding  108s (x1321 over 5h31m)  persistentvolume-controller  no persistent volumes available for this claim and no storage class is set

But as you can see I set presistentSpec enabled property to false.

I try to ask my question on SO

[soroshsabz]

Hi,

That seems very strange. Would you be able to run the log_collector.py script in order to generate a diagnostic package and open a Support Ticket with Redis so that we can take a look? Also, the memory requested (1Gi) is too small. The nodes will not work. We recommend at least 4Gi for testing. I think 3 Gi "might" work.

Laurent.

Where is log_collector.py?

[laurentdroin]

Hi,

The script is here: https://github.com/RedisLabs/redis-enterprise-k8s-docs/tree/master/log_collector

Laurent.

[laurentdroin]

Is it possible that you had previously created a Redis Enterprise Cluster with the same name before? I am thinking the PVC could be from a previous run.
Can you check if the PVC is older than the REC by comparing their creation timestamp?

Laurent.

[soroshsabz]

@laurentdroin thanks you correct