Veeam Remote DR Which Option is best

[zoSMB]

I am currently using veeam for backups of my virtual machines but want to expand and replicate virtual machines to colo site for DR. I have two esx host locally attached to Equalogix SAN and I would like to replicate the virtual machines and host if necessary using host based replication to the DR site that will consist of the same setup of two esx host and equalogix san. What are my options in terms of Veeam replication to replicate data to colo location and bring the hot site online as fast as possible. I am currently at vsphere 4 U1

Also in this scenario and from reading other post should i really considering having veeam installed and setup and colo location instead of the local site? We will initially sync about 1TB of data then small subsets after that. We will only be replicating about 8 virtual machines to colo site as well which will consist of SQL and Exchange

[zoSMB]

also to add to this we will be using vpn site to site connection from local to colo site in this instance should we use network replication mode or vstorage Api mode? Which is faster to in terms of bringing colo site back online as fast as possible? SAN Dell Equalogix at both locations

[Vitaliy S.]

Hello,

You should definitely consider using vStorage API mode with vSphere to leverage all the benefits from VMware CBT technology. If you're going to install a Veeam backup server on the DR site then you should choose vStorage API - Network mode for your jobs. That should give you a pretty good performance rates on your incremental job runs. Also installing Veeam backup server on the DR site protects you from loosing your Veeam SQL configuration DB if the main site goes down.

However to get the best peformance I would install Veeam backup server on the main site and configured all the jobs with vStorage API - SAN mode.

You may find all the information on how to configure a new Veeam Backup Server for SAN access in our FAQ topic

Thank you!

[larry]

I just finished setting up and DR testing a setup like yours. I replicate 600 gigs over a 100 meg pipe with one job. First pass took 12 hours, now each night it is only 40 minutes. I also do a 700 gig backup, about the same 12 hours first then just 1hour a night. Over the same WAN I backup the other direction about 200 gig.
My setup is two ESX setups in both location, both do production, each site is the DR for the other site as well as the off site storage location. I added a spare nic to each ESX server which goes to a isolated DR network( just a standalone switch ) in VC I setup the spare to have the same name as my production network. I have a Veeam server at each site. For DR it is important to have the Veeam server in the remote location running the jobs.
Now the good part. To do a DR test you just open the remote Veeam server and click replicas and failover to the test date. During a real DR you can open up the DR VC and just power on all the replicas, go to the DR Veeam server and disable all the DR replica jobs.
I DR tested a domain controller, 2 SQL servers, 2 Windows 2008 servers in just 15 minutes. I connect a couple of PCs to my isolated DR switch and that was all there was to it. To undo you just fail back. In two hours I did what took a whole day and 4 tech’s before.
If this was a real DR I can just reconfigure my routers, jump the isolated network to the production DR network, I am done.
Didn’t mean to make it so long but hope it helps.

Also now I backup the Veeam servers to tape, they contain the remote data then store the tape locally. No more sending tapes.