chore: update the group of the BackupPolicyTemplate to "dataprotection"

[wangyelei]

No description provided.

[codecov]

Codecov Report

Attention: Patch coverage is 69.57831% with 101 lines in your changes missing coverage. Please review.

Project coverage is 61.63%. Comparing base (038fc44) to head (6632bcd).
Report is 5 commits behind head on main.

Files with missing lines	Patch %	Lines
...trollers/apps/transformer_cluster_backup_policy.go	72.35%	45 Missing and 15 partials ⚠️
.../dataprotection/backuppolicytemplate_controller.go	63.20%	30 Missing and 9 partials ⚠️
controllers/apps/transformer_component_rbac.go	50.00%	0 Missing and 1 partial ⚠️
pkg/operations/custom.go	0.00%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #8203      +/-   ##
==========================================
+ Coverage   61.24%   61.63%   +0.38%     
==========================================
  Files         348      348              
  Lines       40184    40615     +431     
==========================================
+ Hits        24611    25033     +422     
- Misses      13333    13367      +34     
+ Partials     2240     2215      -25     

Flag	Coverage Δ
unittests	61.63% <69.57%> (+0.38%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[zjx20]

Can we move all the logic into BackupPolicyTemplateReconciler?

[wangyelei]

the backupPolicy and backupScheduled are reconciled by the cluster.spec.backup and componentSpecs/shardingSpecs. if move it to BackupPolicyTemplateReconciler, bpt controller will watch the changes of the cluster and handle the logic for all related clusters .

[leon-inf]

The backup policy needs a mechanism or the API to bind to the cluster in a loosely coupled way. For example, referring to Prometheus, the backup policy can define an annotation and well-defined value, so that the cluster only needs to serialize the necessary backup information to this annotation in the defined format, and the backup policy controller can generate the object for it automatically.

[wangyelei]

When using a BPT (Backup Policy Template) controller, every time there is a change in the cluster, the corresponding BPT will reconcile all related clusters to synchronize the backup policies and schedules. If there are many clusters, this can keep the controller very busy, potentially causing blocks and delaying reconciliations. Prometheus, on the other hand, periodically pulls all metrics with the specified annotations, which is a reasonable approach.

[zjx20]

Please resolve conflicts.