Server Fault is a question and answer site for system and network administrators. It's 100% free, no registration required.

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

We have an older cluster using 2 Dell 6850's with 64 gigs of RAM.

We are going to be upgrading the cluster to use 2 Dell R900's with 128 gigs of RAM.

However, the plan to upgrade is going to involve having all 4 nodes in the cluster at the same time.

What I can not find is any documentation about running a cluster with dissimilar hardware. Is this even supported? My guess is that it's definitely not a best practice, but my greater concern is if it's even a supported configurtion.

share|improve this question
up vote 0 down vote accepted

I don't think that is supported, but it should work while you transition everything to the new systems and remove the old nodes; you just need to be careful about storage and network configuration, as usual.

BTW, what kind of application(s) are you running on that cluster? If it runs anything different from standard O.S. services (i.e. file and print sharing), this can get quite difficult; f.e., you must be really careful if SQL Server or Exchange are involved.

share|improve this answer
It's a SQL box. – Jim March Aug 25 '09 at 18:19
Which version? This is how to do it with 2005: – Massimo Aug 26 '09 at 3:09

Yes it is supported without issue. Simply add the new nodes to the cluster, install whatever software is needed on them. Then add the new nodes as available nodes on the resource group(s) in question.

Then fail the resource groups to the new nodes.

Then remove the old nodes.

share|improve this answer
This is 'sorta' the way we are going to do it. Except that the old nodes will remain in the cluster for at least a month. The reason being is that management wants them in the cluster as a fail-back mechanism. This way if the application for whatever reason fails, they can't point the finger at the new servers. – Jim March Aug 24 '09 at 18:37
Shouldn't be a problem. You should be able to put as much time as needed between the steps. – mrdenny Aug 24 '09 at 18:58

Having just gone through this with Server 2008, they make it very clear in 2008 that anything but identical hardware is not supported. 2003 is not 2008, but it does show how Microsoft thinks things should be done.

share|improve this answer
Actually, in Windows 2008 the only requirement is that you pass the cluster validation wizard. Which I believe will allow for small differences in hardware.... – Jim March Aug 24 '09 at 18:36
Jim is correct, as long as the cluster passes the cluster validation wizard it is supported by Microsoft. I have carried this out on very disparate hardware (Dell R900 and 1950 in a cluster == Supported) – Dan Mar 15 '10 at 15:02

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.