Regardless of what type of restore you are running, it’s imperative to ensure that it works, no matter what. Some vendors do a checksum check while others do boot-ability checks. However, nothing compares to a manual check. In this blog, we’ll focus on how to successfully confirm a restore.
Items to check:
Verify that your operating system starts up.
Log in to the machine and ensure all volumes show data.
Ensure services start up.
Check any other specific applications.
Start up any other servers required for a disaster recovery to ensure communication and functionality are good (like a Domain and Exchange server).
NOTE: Make sure to start them all up with no local LAN connectivity and no Internet or you may just cause a disaster yourself. If you could consistently make this a more simple and successful process, wouldn’t you want to?
Data Safe Continuity does that and more.
Every Data Safe backup goes through the following:
VSS Verification - to see if there are issues with how a production machine tracks incremental change.
Filesystem Verification - to check the integrity of the filesystem structure.
Volume Verification - to make sure that all volumes that should be protected are actually backing up properly.
Hardware Independent Restore - to make sure that the information backed up can be restored to other hardware/virtual profiles.
Ransomware Detection - to see if the recovery point shows the likelihood of having ransomware and allow you to take action accordingly.
At least daily, and as frequently as you want:
Boot Verification (or Screenshot Verification) - to make sure you can actually start up the machine in a disaster
Volume Verification - to make sure that all volumes that should be in the virtualization actually show up with data
System and Application Verification - to make sure regular software installed on the production machine starts up when the virtualized
Custom Script Verification - for those of you who want to have specific tests while booted or check for bespoke programs
At a minimum, you’ll get nine automatic checks. If any of these checks fail you can take action BEFORE a disaster happens because making sure that recovering and restoring from a disaster happens as smoothly as possible.
This never gets rid of a scheduled test to make sure the business continuity plan is solid but it will allow you to schedule one per month or quarter to make sure everything works as it should instead of having to hire an intern to just do restores all the time.
With whatever you use to deliver business continuity to your end users, you need to test, but limit the manual work to focus on pressing issues and schedule your disaster tests at your convenience.