Backup & Disaster Recovery
Backup coverage and verification, appliance health, spares, and tested recovery
Every organisation has backups and far fewer have restores. The distance between those two words is where most disaster recovery plans fail, and it only becomes visible on the worst day of the year.
This library treats a backup as unproven until something has been recovered from it, and a recovery target as meaningless until the business has agreed to it.
What this library checks
- Third-party backup of the Microsoft 365 tenant: user coverage, long retention, recent successful runs, and an alert queue somebody has cleared
- Desktop, documents and pictures folders redirected by policy, so what a user leaves on the desktop is captured too
- Image-based backup for endpoints holding critical data, and a documented list of which endpoints those are
- Server storage resilience, volume shadow copies on shared volumes, and directory recovery credentials somebody can find
- Cloud virtual machine backups running to the retention and frequency that were promised
- On-premise backup appliance health: agent coverage of mission-critical servers, local backup success, verified recoverability, cloud replication, capacity headroom
- Failover cluster node and shared storage health, with a clean cluster event log
- Cold, warm or hot spares held for each type of router, switch and access point in use, stored where the outage will be
- Manual fallback procedures for when a point-of-sale or line of business system is offline
- Recovery time and recovery point objectives defined and agreed with the business, and a failover test that proves systems come back inside them
- Copies held in a geographically separate location, so a site-level event does not take the recovery with it
- Configuration backups for network and surveillance infrastructure, not only data
- An incident response plan that is documented, approved and signed
Evidence you will be asked for
Every check in Spectra Core carries its own evidence. These are the artefacts an auditor, insurer or board most often wants to see for this area.
- Backup success history across the review period, with failures and how each was resolved
- Record of the last failover test: date, scope, and whether recovery landed inside the agreed objective
- Written recovery time and recovery point objectives, agreed by the business
- Documented list of endpoints holding critical company data
Where this usually goes wrong
Downtime tolerance is a leadership decision, not a technical one. IT can build to any target, but only the business can say which target it is paying for, and that conversation almost never happens until after an outage has already answered the question.
Run this library against your environment
Every check comes with a risk level, a review frequency, an assignee and a plain-language rationale, so the output reads for leadership as well as for engineers. See the full workflow, or read how internal IT teams and managed service providers use it.
Related libraries
Server Rooms
Cooling, cleanliness, cable labelling, lockable racks and photographic evidence
Power Management
UPS runtime, managed shutdown, generator availability and supply configuration
Microsoft 365 Security
Tenant MFA, admin accounts, Intune policy, Teams permissions and app registrations
Remote Monitoring & Management
Agent currency and device naming across the managed estate
Last updated