When the Cloudera Manager UI loses connection with CDP Private Cloud Data Services on OpenShift, how can we operate the cloud data service?

[#1]

Hi everyone, I would like some clarification. Thank you.

We are using CDP Private Cloud Data Services 1.5.2 on OpenShift 4.12.46 (Kubernetes 1.25).

Currently, we are unable to log in to the OpenShift Container Platform (OCP) using 'oc login'. Therefore, we are using the 'localhost-recovery.kubeconfig' file to execute OCP commands.

Source: https://access.redhat.com/solutions/4845381

To use the 'localhost-recovery.kubeconfig' file, run the following commands:

export KUBECONFIG=/root/localhost-recovery.kubeconfig
oc get ns

We deleted the CDP Private Cloud Data Services on OpenShift from the Cluster Manager (CM) UI. As a result, the UI no longer displays the Private Cloud Data Services, although the actual Private Cloud Data Services still exist on OpenShift. However, we cannot perform operations through the CM UI (as it is not visible in the UI).

In this situation, we have a few questions:

• How can we remove the Private Cloud Data Services from OCP?
• Is there a possibility to see the existing Private Cloud Data Services cluster again on the CM UI?

[#2]

Hello @steinsgate 

Thank you for bringing this

Since the cluster has been removed from the Cloudera Manager the only way I think would be to re-install

Also have you followed the below document for uninstalling the cluster

https://docs.cloudera.com/cdp-private-cloud-data-services/1.5.2/installation/topics/cdppvc-installat...

[#3]

Hello @upadhyayk04 

Thank you for the reply.
Yes, we followed the documentation to delete the cluster.
Then we encountered the situation as shown in the screenshot above.
Even though the WEB UI indicates that it has been deleted, checking the logs or going to OpenShift reveals that it has not been deleted yet.
I would like some clarification. Thank you.
So, we should wait for the OCP cluster to return to normal login before reinstalling the cloud data service in the same namespace.
The recovery.kubeconfig we are currently using is invalid.

oc config view --minify=true --flatten=true > kubeconfig0403.yaml

[#4 · Accepted solution]

Hello @steinsgate 

In case you want to completely remove the cluster. I would say we can think of the removing the namespace from the Openshift all together and then start fresh