Metrics agent ConfigMap should use cluster name override

[olim7t]

When creating a cluster with a name override:

apiVersion: k8ssandra.io/v1alpha1
kind: K8ssandraCluster
metadata:
  namespace: test
  name: demo
spec:
  cassandra:
    serverVersion: 4.1.0
    clusterName: override # <<< name override here
    datacenters:
      - metadata:
          name: dc1
        size: 1
        racks:
          - name: rack1
        storageConfig:
          cassandraDataVolumeClaimSpec:
            storageClassName: standard
            accessModes:
              - ReadWriteOnce
            resources:
              requests:
                storage: 100Mi

The metrics agent ConfigMap is named demo-dc1-metrics-agent-config.
This is inconsistent with other Kubernetes resources (pods, etc) which are named after the override.

This line should use c.Kluster.SanitizedName(). Additionally, it should check that the concatenated name doesn't have more than 253 characters.

We should also audit other generated names to make sure that those rules are consistently applied.

[olim7t]

Note: this ticket used to cover invalid DC names too, but that is a more widespread issue so I moved it to a dedicated ticket: #1141.