Most backup conversations begin with a straightforward question: can we restore it?
That was how I approached recovery for years—protect the systems, maintain the retention schedule, replicate where appropriate, and make sure we could recover after a failure. At VMware Explore, I kept coming back to a harder question:
How do you know your backups are good?
For cyber recovery, recoverable is not the same as trustworthy. The newest recovery point may already contain attacker persistence, compromised credentials, damaged data, or other changes you do not want to return to production.
A backup can be completely recoverable and still be the wrong thing to restore.
Recovery starts looking more like an investigation: identify candidate recovery points, isolate them, validate the systems and data, and build enough confidence before allowing anything back into production.
I wrote more about that shift here:
Cyber Recovery: Just Because You Can Restore It Doesn't Mean You Should
How does your organization decide that a recovery point is trustworthy enough? What validation happens before recovered workloads are allowed back into production?
That was how I approached recovery for years—protect the systems, maintain the retention schedule, replicate where appropriate, and make sure we could recover after a failure. At VMware Explore, I kept coming back to a harder question:
How do you know your backups are good?
For cyber recovery, recoverable is not the same as trustworthy. The newest recovery point may already contain attacker persistence, compromised credentials, damaged data, or other changes you do not want to return to production.
A backup can be completely recoverable and still be the wrong thing to restore.
Recovery starts looking more like an investigation: identify candidate recovery points, isolate them, validate the systems and data, and build enough confidence before allowing anything back into production.
I wrote more about that shift here:
Cyber Recovery: Just Because You Can Restore It Doesn't Mean You Should
How does your organization decide that a recovery point is trustworthy enough? What validation happens before recovered workloads are allowed back into production?