WHAT

Store the reserved  kafka in the database with a special marker (a new database "deployment_type" column which can be "real" or "reserved"), assign them to each data plane cluster - using the cluster_id info. **

WHY

To track the managed kafka in the database so that they follow a proper deletion flow: from the data plane first and then from the fleet manager database

HOW

- A new database "deployment_type" column which can be "real" or "reserved". The default value for this column has to be "real".

• Make sure that the reserved kafka are reconciled in the database each time for each cluster. 

• When more reservations are needed i.e the reservation configuration increases make, we have to increase the reserved kafkas for each data plane cluster in the database
• When the reservation configuration value decreases, the reconciler has to reduce the reserved kafkas in the database. For this, the reconciler will have to mark the first N reservations as deprovision: Where N is the difference between the old reservation and the new one
• Send a warning alert to Sentry when we try to reserve more kafkas than the maximum number of streaming units supported by the cluster. This will enable the Control plane to re-adjust the reservation. This is a nice to have

The capacity reservation configuration will be exposed via https://issues.redhat.com/browse/MGDSTRM-8736

DONE

Include the following where applicable:

reserved managed kafkas are stored in the database for each data plane cluster and they are continuosly reconciled