Curious how MSPs and admins here handle this. If the attacker may have been inside for weeks, how do you decide which restore point to bring a VM back from? The latest backup might already be compromised, and going too far back loses a lot of data. Do you test the restored VM in isolation first, and what do you check before you call it the recommended recovery point?
Disclosure: I work on CyberSnap, a ransomwar
I’d anchor the restore point to the earliest confirmed attacker activity, not the latest backup: EDR alerts, unusual logons, new admin accounts, backup catalog changes. If that timeline is incomplete, use the oldest immutable/offline backup that predates symptoms and accept some data loss. Restore to an isolated VLAN, boot without production credentials, scan, check startup tasks/services and AD/cloud persistence, then rotate credentials before reconnecting.
That order makes sense, thanks. One thing I’d add is retention, since if backups only go back a couple of weeks you may not have anything that predates a long dwell time. I’d also write the restore order into the runbook ahead of time, so nobody is deciding under pressure which system comes back first. Disclosure: I work on CyberSnap, a ransomware recovery tool.
After a ransomware attack, I would choose a restore point from before the infection started. First, I would make sure the backup is clean and free from malware. I would also check that the system is secure before restoring the data to avoid another attack.