Disaster recovery option: Veeam VS SRM

One thing to note is that Veeam doesn’t care about your storage in most cases but does have some storage integrations depending on the type of arrays you’re using and if you’re performing any storage-based replication.  If you’re performing your replication of VM’s at the VM/Host/vCenter level then it doesn’t matter there for SRM.

But the one gotcha I learned is that if you’re performing storage-based replication, your storage must match on both ends.  SRM uses SRA’s (Storage Replication Adapters - basically a driver written by the storage vendor), and SRM requires (or at least, but I would guess still does) that the SRA is the same on both ends, meaning the storage needs to be the same on both ends. 

It’s been a few years, but my use case was that I had a Dell SC (Compellent) in the primary datacenter and was replicating to a Dell PS (Equallogic) on the recovery datacenter.  Those use different SRA’s and thusly wasn’t able to use SRM.  The solution I presented was to upgrade their VMware licensing so that we could perform VM-based replication, but the client instead chose to replace the Equallogic with another Compellent and it worked great after that.  Had this client been using Veeam (instead of Quest RapidRecovery), then it would have been a non-issue.

May not apply to everyone, but just a tid-bit that I learned and share whenever I can.

Yes, I remember the SRAs being a pain if they did not match on both sides.  Luckily, I was testing Nimble, so they had the proper SRA, and it was in both cluster we used SRM.

In my case, I was just getting ready to enable things and just happened to come across a single line about it in one of the guides/manuals.  It wasn’t glaringly obvious.  Opened a case up with VMware right aware and they confirmed that the SRA’s must match.