Velero data movement - data path design

[Lyndon-Li]

We need to create a design to answer below data movement data path questions for backup:

1. After a snapshot is taken, how is the snapshot exposed to the data mover pod
2. How many snapshots are exposed to one data mover pod, what is the parallel processing like
3. For file system data movement, after the snapshots are exposed, how to launch Kopia uploader as mentioned in the design Unified Repository Design #4926
4. For file system data movement, how is the data organized in the Unified Repository
5. What about block data movement, we can set it as non-goal and cover it in a separate design in future. But at present, we at least need to clarify how to choose the data movement mode(file system, block) and how is the mode persisted
6. How does the uploader writes data to Unified Repository. This should be the same with pod volume backup as shown in Unified Repository Design #4926, but we need to clarify the entire data flow
7. How is the Unified Repository is invoked for initialization

Similarly, we need to answer some questions for restore:

1. After the target PV is created, how is the PV exposed to the data mover pod
2. How many target PVs are exposed to one data mover pod, what is the parallel processing like
3. For file system data movement, after the PVs are exposed, how to launch Kopia uploader
4. An entire workflow to show how the uploader reads data from Unified Repository

[reasonerjt]

This will be covered in v1.11 when we deliver the overall design of data mover in velero.
Closing as dup of #4112