Reconcile Loop Fails to re-reconcile the group in case the group was deleted from Backend

[vinamra28]

Actual Behaviour

Once a group CR is applied on the cluster, it onboards the same across backends specified. If an admin of that respective SaaS backend goes and drops that Group or User from a backend, Usernaut isn't able to re-reconcile that and goes into error state. This happens because of the cache layer that acts as an intermediate representation layer of all the backends and this cache stores the IDs provided by each backend.

Expected Behaviour

In such scenarios, Usernaut should be able to self-heal the system and recreate the respective user and team on the backend.

Steps to Reproduce

1. Create a group CR on the cluster

   apiVersion: operator.dataverse.redhat.com/v1alpha1
   kind: Group
   metadata:
     labels:
       app.kubernetes.io/name: usernaut
       app.kubernetes.io/managed-by: kustomize
     name: dataverse-platform-usernauttest
     namespace: usernaut
   spec:
     group_name: dataverse-platform-usernauttest
     members:
       users:
         - user1
         - user2
     backends
       - name: fivetran
         type: "fivetran"

2. Go to fivetran and drop the team

3. Make changes in the user list of the applied CR and re-apply the same on the cluster

   apiVersion: operator.dataverse.redhat.com/v1alpha1
   kind: Group
   metadata:
     labels:
       app.kubernetes.io/name: usernaut
       app.kubernetes.io/managed-by: kustomize
     name: dataverse-platform-usernauttest
     namespace: usernaut
   spec:
     group_name: dataverse-platform-usernauttest
     members:
       users:
         - user1
         - user2
         - user3
     backends
       - name: fivetran
         type: "fivetran"

4. Notice the behaviour.