EKS collector reports cluster as permanently deactivated

Products

• CloudHealth

Issue/Introduction

When redeploying a collector on a cluster, you encounter the following error message:

ERROR -- : Conflict [409]: Failed to post cluster state to https://containers-api.edge.cloudhealthtech.com/v2/containers/kubernetes/state?cluster_id=XXX&sample_time=XXX&k8s_object_type=XXX&k8s_agent_version=XXX&cluster_uid=XXX&auth_token=API_TOKEN_REDACTED. Error: {"messages":"Unable to collect. Cluster XXX (UID: XXX) has been permanently deactivated."}

Resolution

This error message appears when attempting to redeploy a collector using the same cluster name as a previously deactivated cluster that does not have a UID in the platform. That is because during redeployment, the platform assigns the first detected UID to the existing deactivated cluster instead of the new one.

To resolve this issue, redeploy the collector to the cluster using Helm, but set the CHT_CLUSTER_NAME variable to a new value, such as CHT_CLUSTER_NAME-new.

Keep in mind that the cluster itself does not need to be renamed; it's only the CHT_CLUSTER_NAME variable.

The screenshot below shows a deactivated cluster with no UID in the platform.