Prometheus metrics for Fusion Backup and Restore Service operations

Fusion currently exposes a limited set of Prometheus metrics for backup and restore operations. While sufficient for basic monitoring, these metrics do not provide enough detail for production-grade observability.

Problem Statement:
The current Prometheus metrics do not provide sufficient visibility into job status, execution trends, restore accountability, and storage usage, resulting in operational blind spots.

Proposed Metrics:
1. Backup Job Metrics
fusion_backup_job_status{job_id, namespace, workload, status}
- status: success | failed | running | canceled

fusion_backup_job_duration_seconds{job_id, namespace, workload}
fusion_backup_job_start_timestamp{job_id}
fusion_backup_job_end_timestamp{job_id}
fusion_backup_objects_total{job_id}
fusion_backup_data_bytes_total{job_id}

2. Backup Summary Metrics
fusion_backup_summary_last_status{namespace, workload}
fusion_backup_last_success_timestamp{namespace, workload}
fusion_backup_failure_total{namespace, workload, reason}

3. Restore Job Metrics
fusion_restore_job_status{restore_id, namespace, target, status}
fusion_restore_job_duration_seconds{restore_id}
fusion_restore_job_start_timestamp{restore_id}
fusion_restore_job_end_timestamp{restore_id}
fusion_restore_triggered_by{restore_id, user, auth_type}
auth_type examples:
user
service_account
automation

4. Restore Audit Metrics
fusion_restore_total{namespace, initiated_by}
fusion_restore_failure_total{namespace, reason}

5. Backup Storage & Capacity Metrics
fusion_backup_bucket_usage_bytes{bucket_name}
fusion_backup_bucket_objects_total{bucket_name}
fusion_backup_retained_snapshots_total{namespace}
fusion_backup_storage_growth_bytes{bucket_name}

Criteria:

• Metrics are exposed via the existing Prometheus endpoint
• Metrics follow Prometheus naming and labeling best practices