rook: Dashboard: data on object-storage throws error.

Is this a bug report or feature request?

  • Bug Report

Deviation from expected behavior: I have applied the Helm charts for rook-ceph and rook-ceph-cluster, respectively. Afterwards I apply yaml for cephObjectStore, cephObjectStoreUser, obc, sc to get object-storage. It is intentional not to inject values with respect to object storage into the valuefile to the charts as it is desired to get a user and bucket with unique credentials.


apiVersion: ceph.rook.io/v1
kind: CephObjectStore
metadata:
  name: my-store-1
  namespace: rook-ceph
spec:
  dataPool:
    failureDomain: host
    replicated:
      size: 1
  gateway:
    instances: 2
    port: 80
  metadataPool:
    failureDomain: host
    replicated:
      size: 1

--- 
apiVersion: storage.k8s.io/v1
kind: StorageClass
metadata:
  name: my-store-1-storageclass
parameters:
  objectStoreName: my-store-1
  objectStoreNamespace: rook-ceph
  region: us-east-1
provisioner: rook-ceph.ceph.rook.io/bucket
reclaimPolicy: Delete

--- 
apiVersion: ceph.rook.io/v1
kind: CephObjectStoreUser
metadata:
  name: my-store-1-user
  namespace: rook-ceph
spec:
  displayName: my-store-1-display-name
  quotas:
    maxBuckets: 100
    maxObjects: 10000
    maxSize: 10G
  store: my-store-1

--- 
apiVersion: objectbucket.io/v1alpha1
kind: ObjectBucketClaim
metadata:
  name: my-store-1-my-bucket-1
  namespace: rook-ceph
spec:
  additionalConfig:
    maxObjects: "10"
  bucketName: my-bucket-1
  storageClassName: my-store-1-storageclass

The dashboard is exposed using a virtualService, with the values to the chart:

cephClusterSpec:
  dashboard:
    enabled: true
    port: 7000
    ssl: false

When entering the dashboard the UI is nicely presented and it shows that it has noticed the 2 rgw-instances configured as expected, but when entering the page about object storage it throws an error in the mgr-log about no route to host at host=‘192.168.200.11’, port=8080 as shown in the logs. This makes the UI say: “Error connecting to Object Gateway. Please consult the documentation on how to configure and enable the Object Gateway management functionality.”

Expected behavior: The dashboard is enabled to work with object storage without throwing errors such that data can be rendered in the dashboard.

How to reproduce it (minimal and precise):

  • Have a running ceph-cluster based on chart for version 1.7.8
  • Apply the yaml files and wait for RGW pods to start
  • Enter the dashboard and click on “Object Gateway”
  • See the error being thrown in the logs for the mgr-pod

Environment:

  • OS: Centos8
  • Rook version: 1.7.8
  • Ceph version: 16.2.6
  • Kubernetes version: 1.21.3
  • ceph status: ‘health_ok’

Logs:

About this issue

  • Original URL
  • State: closed
  • Created 5 years ago
  • Comments: 15 (7 by maintainers)

Most upvoted comments

Hmm. None of the above does anything that helps. But in general, I believe that your are right in that there is “something” that blocks it - it does not seem like it can be anything else. Based on our discussion here we also tried to deploy it in a Azure environment where we saw that it actually did work - so apparently it all works as expected. I will look into what might be wrong with my environment but as of now it is no longer an issue related to Rook Ceph.

But thanks for all the assistance - much appreciated.

@jmolmo can you take a look?