Data Protection Operations, Monitoring, and Management
Domain 7 tests day-to-day operations: monitoring jobs, managing capacity, handling alerts, reporting SLA compliance, and troubleshooting. Expect scenario questions about a failed backup, a full protection-storage system, or a missed RPO.
Core rules
- Job monitoring checks success, duration, throughput, and error codes. A green job can still be unusable if the catalog is inconsistent or the restore fails.
- SLA reporting compares actual RPO/RTO and success rates with targets. Report by asset group, application, and site, not just a global percentage.
- Capacity management tracks usable capacity, dedupe ratio, compression, growth rate, and retention changes. A rising dedupe ratio may hide a growing full-backup footprint.
- Alerting should be actionable: failed jobs, missed schedules, replication lag, capacity thresholds, certificate expiry, and failed logins. Alert fatigue is a design defect.
- Lifecycle management covers software upgrades, firmware compatibility, certificate rotation, and end-of-support dates. Test upgrades in a lab.
- Troubleshooting starts with scope: one asset or all assets, one policy or all policies, one site or both. Check credentials, network, proxy, storage capacity, catalog, and the client log bundle.
- Runbooks document recovery steps, contacts, dependencies, and validation. A runbook that has never been tested is a hypothesis.
Worked example
A daily backup fails for 30 SQL databases but succeeds for VMs. The scope points to the database policy or the SQL host, not the DD array. The operator checks the PowerProtect Data Manager job log for authentication errors, confirms the service account password has not expired, verifies the SQL writer and VSS status, and checks whether the protection policy changed. After fixing the account, a restore of one database validates the fix. The incident is logged, and an alert is added for service-account expiry.
Traps that cost marks
- Monitoring only job status.
- Reporting a global success rate that hides a failed application.
- Ignoring capacity growth until the array is full.
- Changing retention without checking capacity impact.
- Skipping restore validation after a fix.
- Treating every alert as equally urgent.