Kahu: Requirement analysis for Job based data protection framework

[AmitRoushan]

Issue/Feature Description:
KAHU project under the soda-cdm organisation deals with data protection aspects of container data management on Kubernetes .
Currently Kahu supports different volume backup and backup location with Provider framework. The framework provides a plugin approach to integrated Volume/BackupLocation driver in runtime itself.

Currently Volume/BackupLocation drivers are deployed either as “Deployment” or “Statefulset”. The solution works perfectly if the drive uses APIs to connect with respective servers. But it would not work

• If backup/restore drivers are data movers (like Restic).
• If volumes are topologically distributed
• Currently Kubernetes does not support hot mount functionality, So if any Volume backup driver needs “Volume” mount for backup/restore. Current approach will be failure

Why this issue to fixed / feature is needed(give scenarios or use cases):
Data protection is a scheduled operation. So running a driver process and waiting for a backup/restore request to process is a waste of memory and CPU cycles.

An alternative approach is to make accept volume backup driver and backup location driver as Pod template and schedule a Kubernetes job to process backup/restore request

The issue is submitted to discuss an alternate approach (Job based data protection) to handle data protection.

How to reproduce, in case of a bug:
NA
Other Notes / Environment Information: (Please give the env information, log link or any useful information for this issue)

Initial draft is compiled here