Run backup agents as part of the FDB cluster

Currently, backup agents are provisioned separately from other processes (e.g., master, proxy, tlogs) of FDB. This separation often confuses user -- they submit a backup request to the FDB, but see no progress being made because there are no backup agents running.

The new backup system introduces "Backup Workers" inside FDB cluster, which uploads mutation logs to blob storage. However, backup agents are still required to save database snapshots. It seems reasonable to consolidate backup agents processes inside FDB cluster so that:

- Operators do not need to separately provision backup agents;
- The total resource requirements for backup can be planned in one place.

Start by reviewing how backup agents are currently provisioned separately from the FoundationDB cluster and how the new Backup Workers operate inside it. Done means backup agents are consolidated into the cluster so operators can provision them there and plan total backup resources in one place.

Written by the indexing model from the issue text.