Fixed autoRebalance status not reset when auto-rebalancing or Cruise Control is not enabled

[ppatierno]

Type of change

• Bugfix

Description

When the cluster starts without Cruise Control enabled or without auto-rebalance configured, the corresponding status is set anyway (which on cluster creation is anyway wrong, because it's not a scaling up operation).
When Cruise Control or auto-rebalance configuration is removed, the status is not reset as well.
For example, you create a Kafka cluster with:

cruiseControl: {}

on creation, the status contains the following:

status:
  autoRebalance:
    modes:
    - brokers:
      - 3
      - 4
      - 5
      mode: add-brokers

It hasn't to be there, because cruiseControl.autoRebalance is not enabled at all and it's even cluster creation not scaling up.
No auto-rebalance runs but the status is anyway wrong.
Another example, you create a Kafka cluster with cruiseControl.autoRebalance configured so the status is:

status:
  autoRebalance:
    lastTransitionTime: "2024-09-28T14:14:45.485639784Z"
    state: Idle

and then you remove the autoRebalance part of it or even the entire cruiseControl section.
The status.autoRebalance should disappear because keeping the info about the "Idle" state doesn't make any sense anymore (auto-rebalance is not configured).

This PR fixes the above issue(s), also improving the way the KafkaAssemblyOperator can make the decision about running or not the auto-rebalacing reconciliation at all. This way some state initialization is also pulled out from the reconcile of the KafkaAutoRebalancingReconciler.

Checklist

• Write tests
• Make sure all tests pass
• Try your changes from Pod inside your Kubernetes and OpenShift cluster, not just locally

[ppatierno]

/azp run regression

[azure-pipelines]

Azure Pipelines successfully started running 1 pipeline(s).

[scholzj]

One nit. LGTM otherwise.

[ppatierno]

/azp run regression

[azure-pipelines]

Azure Pipelines successfully started running 1 pipeline(s).