Force Sync VPG question

Hi Tom,

Many years ago I discovered the same issue as you. A server would blue screen during a test failover. Turns out it was because at some point in history, a snapshot was taken before a software upgrade, the upgrade went south, so the machine was rolled back to that snapshot.

The issue here is that Zerto has no visibility into the snapshot rollback. It has no idea that the blocks that it saw as written (and replicated those blocks) have been undone. That creates a condition where your VMDK files are forever out of sync. A force sync fixes this issue because it compares every block, so that fixed it for me.

This could be what happened to you, or it could of been something else, but I thought it was worth mentioning my cause, in the event it could be the same as yours, and if not, it’s still worth noting as I feel it’s a big hole right now that people might not even know about.

Ok so the reasoning for telling that whole story is, like you, I needed a way to automate a force sync. I figure I could script a way to monitor vCenter for any snapshot revert tasks (which I did) then if it detects that, execute a force sync on that VPG, which I was able to do using the Zerto commandlets.

So to answer you question, yes, it’s entirely possible to script a force sync, then have a scheduled task to execute it. Just be aware, a force sync can take some time, during which time you are not logging new checkpoints so if you have a real DR event in the middle of a force-sync, you will have data loss.

Hope that helps.

Then in short, the answer is yes, it can be done.

Matthew, can you share your script to force resync?

Hi Enrique. As of version 6.5 Zerto automatically detects this condition, so a manual script is no longer necessary.