Exchange 2003 Failover Best Practices


We currently have a production and a failover site that each have their own Exchange 2000 server. Each sites Exchange Server is connected to a SAN which replicates data between the two sites. The sites are currently connected by a VPN with about 1MB of bandwidth, but will soon be connected with 40-45MB of bandwidth. We are in the process of installing an Exchange 2003 Cluster at our production site and migrating all of our users to that. We want to install an Exchange 2003 server at our failover site but I'd like some ideas and best practices on how to have the failover site server configured for best/most uptime incase of a disaster.

Some things I've read say that it has to be a cold standby server that gets booted up and the database restored back onto it if the production site becomes unavailable. It seems like there would be an automatic or "hot" way to do this.

Is there?

Who is Participating?
Steve McCarthy, MCSE, MCSA, MCP x8, Network+, i-Net+, A+, CIWA, CCNA, FDLE FCIC, HIPAA Security OfficerConnect With a Mentor IT Consultant, Network Engineer, Windows Network Administrator, VMware AdministratorCommented:
Exchange is still a very funny animal.  You can Cluster or you can use multiple servers so that if one goes down it does not affect everyone.  Unfortunately the recommended recovery method for the non clustered environment basically boils down to a server sitting in cold standby in case of disaster.  There is no automatic or Hot Swap capability.
David WilhoitConnect With a Mentor Senior Consultant, ExchangeCommented:
true, no built-in capability. But if your SAN is already replicating data to your hotsite, then see if you can leverage that by replicating the Exchange data to a cluster at your hot site. There are several apps that can do this for you, like DoubleTake or WANSync HA from XOSoft. Replistore from EMC is used on their SANs. IT really depends on your service level agreement; how important is email? How long is management willing to wait in the case of total disaster, loss of your whole physical site?

How quickly you want it back will determine the cost of the solution, but it sounds like you have most of it already in line.
Is that about right?

joshsfinnAuthor Commented:
Thanks for the input. WANSync looks good but the more I look at this, the more I think that our SAN can do the same thing.

We have identical EMC CX500 SANs in our Main and DR site that repliace data. The Exchange data and logs will go on the SAN.

According to a Microsoft white paper called "Server Clusters: Geographically Dispersed Clusters", I can put the servers in the main site and the server at the DR site in the same VLAN to create geographical cluster.

If I can use the SANs to copy the data, my feeling is that I should be able to do this.

Of course, I've never done anything like this before so I'd like some tips from people who have or know more than me. :)

If this can be done, is my best option to create a geographical cluster with 3 nodes - two at the main site and one at the DR site? Also, will that even work because even though the DR server can be in the same VLAN, it will have a different IP address than the two servers at the main site.

Sorry for the ramble and I hope I'm not making this confusing.. :)

Thanks again!
I am currently looking into a similar solution, the key to making this work is dependent on a few factors:
 there mst be less than 500ms of network latency between the Dr and main sites & as I understand it the Clusters must be part of the same subnet and share a common "heartbeat" subnet as well.
I too have two sans and am in the process of configuring data replication from our main site to the DR, since we have dedicated fiber between the sites I have several options, I have allready extended an existing LAN out to thr DR site without any issues.
In your case if the DR server is on a different subnet, you may have to do some creative routing/NAT'ing in order to make you "geographcally dispersed" Cluster work, it would probably be unsupported from MS if there is an option.
Another option we are looking at is "virtualizing" the email and file servers, then by the data replication we can bring up a replicated environment in llittle or no downtime.

Question has a verified solution.

Are you are experiencing a similar issue? Get a personalized answer when you ask a related question.

Have a better answer? Share it in a comment.

All Courses

From novice to tech pro — start learning today.