So Site Recovery Manager can satisfy all the requirements we're needing to replicate VMs to a seperate physical site? I see it does handle making the process of recovering from a site disaster much more simplified. Thanks!

We found that locally, the SVC worked fine. It is a nice concept, but it is also very cryptic and complicated as to the setup of the environment. Practically requires a class to know how to manage. We had to have IBM come set it up because it was so complicated and nobody locally was certified to set up and configure it. Our biggest deal breaker was with the global replication. It does not buffer if it gets behind and breaks the replication extremely easily. Also requires the use of Fibre-to-Copper converters from Qlogic (SANboxes). We even had issues with it running VMWare ESX 3.5 where it would literally reset VMs whenever the global replication broke. This is an issue we had taken up to the IBM architects as well as VMWare. Had a group of IBM engineers fly in here from Chicago and Texas to try to figure out what the issues were. So, any documentation you see from IBM about now requiring vSphere4 is due to our company's issues for the past 8 months. However, the replication still breaks at the slightest hiccup or if it gets behind. We were one of the first companies in our area to have one of these setups; however, we found out that there is another company around here that is having the exact same issues we are experiencing. So, we looked at a solution from Falconstor (IPStor) which met all our needs and then some. Issue right now is cost and my needing to just come up with various options before we come to a final decision. If you are looking for a virtualized disk environment with easy replication and failover, look into Falconstor. It is cheaper than the SVC I believe and works fantastically. I literally did a full-site failover in a few clicks and was able to throttle replication bandwidth easily. Our replication bandwidth is limited to about 25Mb during the day with up to 40Mb in off-hours. All replications are asynchronous. With Falconstor, we had it set to copy the deltas every hour and it kept up with it just fine. Was extremely efficient in the data handling. Lets you take timestamp whole LUN snapshots as well so if you corrupt a VM you can restore it or other files easily.

Hope this helps! Still wondering if Site Recovery Manager replicates actual VMs or whether it just manages a failover capability.