I've already seen a few discussions touching on this topic of DR solution using DataKeeper to perform block level replication to a DR site hundreds of miles away from the primary data center. Even though they have been helpful, I have a few questions I'm not sure I've found the answer to. So here's the situation: we have an 2-node Active/Active cluster with 2 SQL Server 2008 R2 FCIs on each node, so a total of 4 instances, with SAN storage. We have 200+ databases between these instances ranging from as little as 50MB to as large as 150GB. We have a DR datacenter 650 miles away with a stand alone server for SQL Server DR databases and need to replicate/mirror the data in the primary cluster to the DR server. All three servers (2 nodes in cluster and DR server) are identical: 32-core CPU, 384GB RAM, however, like I mentioned, the DR server is not a node in the cluster. I was looking at Database Mirroring for a solution to have the data replicated to the DR site but there are worker threads limitations that would limit the number of databases to mirror, especially on the mirror server. Also, DB Mirroring only works at the database level and anything outside (server logins, agent jobs, etc) would have to be kept synchronized manually. But the showstopper for me was not being able to mirror all the databases. Then we started looking for other options and SIOS DataKeeper kept coming up as a great and more complete solution as it would mirror EVERYTHING on the drives and do away with the mirrored-databases limit, and I believe would also be easier to administer/maintain.
Question: Do we need to use FusionIO drives for DataKeeper to do it's job? Can we replicate from the SAN currently holding the data to the DR site? I believe it is required if you're setting up a SANless cluster but is it required when you already have a SAN and you just want to replicate that data?
Question: Does the DR server need to be part of the primary failover cluster or can the data be replicated to an independent server?Should the DR server be a part of the cluster for best practices? Why?
Question: Which edition of SIOS DataKeeper would be appropriate for this setup?
I believe I had some other questions but can't remember right now. I will update this post when it comes back! I appreciate your feedback. Thank you!
Edit: I remembered another question. Could I use DataKeeper to replicate data for stand alone instances to the DR server? We do have a handful of 2008 R2 instances that host databases which are either not cluster compatible or need their own server.
Leroy G. Brown