So here's my plan, can you guys shoot holes in this for me?
We currently have a production data center that replicates to a dr data center which is in another building on the same campus (separated by a parking lot).
We also back up limited amounts of production data and ship that off via tape nightly.
Here is what sucks with this right now.
DR and Production are too close for comfort. If we had a tornado come through or a power outage, this could take out both buildings, so both production and DR could be down at the same time.
Tape sucks, takes too long, we can only fit 8 of our 221 VM's onto tape.
Here is what I'm thinking of doing.
Finding a colo facility, put my DR replication hardware at that site, put in a fat pipe (500Gbps), and then just replicate over this wire. Using Veeam One I see that I have about 753GB of changes per day that I replicate. That is on average of 31GB per hour I need to move. Why? Well, if disaster hits, I don't want to be restoring anything from tape. I just want to boot up my DR site and have all systems back online in a short time frame, it just seems much easier and simpler in design vs. some elaborate backups structure (recall tapes, restore tapes, uncompress data, inject data, etc, etc, etc).
Now, I have my too close for comfort issue fixed but I still have a need for backups. I need to be able to recover files that are deleted, corrupted, moved, encrypted by viruses, etc. I'm thinking that I put in a Cisco C3160 and keep that at my production site. I could then have a very fast, local backup target that I could keep all sorts of recovery points on. Nightly for a week? Sure. Weekly for a month? Sure. Monthly for a year? Sure. All of these backups would be on-site but it doesn't matter to me because I have a fully replicated DR site that I would use if catastrophe hit.
One thing to keep in mind is we don't really do archiving here. Once data makes it into our systems, it stays in our systems.
Thoughts, concerns with this strategy?