Utilizing all cluster nodes at the same time in an active passive win2k two node cluster

Hello Everyone,
I'm planning on setting up an active/passive two node MS cluster in a lab environment. So far I have got some directions to perform the basic configuration. I'm planning on installing both Exchange 2003 & SQL 2005 on this cluster. I do understand that I need a SAN or some sort of shared storage such as SCSI storage to properly implement this. Lets assume that I have met all hardware and software requirements. My plan is to utilize both servers at the same time. There are two questions regarding Microsoft clusters,

1. Is it possible to have   Node-1 active with Exchange and node-2 active with MS SQL in this cluster? E.g. let say that node-1 goes down, then node 2 take over the Exchange tasks or If node-2 goes down, then SQL tasks will be transferred to node-1.  If there are any special requirements to achieve this, what are they?

2. Is it possible to setup either an Exchange or MS SQL active/passive cluster between a dedicated hardware server and virtual machine (VM) running on a virtualized environment?  If there are any special requirements to achieve this, what are they?
Who is Participating?
I wear a lot of hats...

"The solutions and answers provided on Experts Exchange have been extremely helpful to me over the last few years. I wear a lot of hats - Developer, Database Administrator, Help Desk, etc., so I know a lot of things but not a lot about one thing. Experts Exchange gives me answers from people who do know a lot about one thing, in a easy to use platform." -Todd S.

Hello rockygold,

1. Yes it is possible to have this type of setup. I dont know if I would recommend it for production environments - especially with SQL and Exchange and they way that they want to take all of the kernel memory - but for a test environment you should be ok.
   The only requirement for this is to have seperate cluster groups for each resource - which is what you should do anyway. This will also require a seperate shared disk. In total, you will need at least 3 disks (1 for quorum, 1 for SQL, 1 for exchange). While there are ways you can do this with SCSI clustering, I would recommend that you use ISCSI or FC SAN clustering for what you are trying to do. Doing this with SCSI clustering would most likely require 2 controllers in each server and 2 DAS units.

2. It is possible to setup a physical / virtual cluster. ISCSI shared storage works great for this. I have not tried this in a prodution environment, nor do I know if this would be a best practice. But that doesnt apply for your situation. The same requirements apply as in question 1. While you may be able to find a way to do this with SCSI shared storage, I wouldnt recommend it even if you could get it to work.

BTW, if you are just using this for testing, there is a great free linux distro that I use for my iscsi shared storage called openfiler (http://www.openfiler.com/). You can setup a spare server as the shared storage and use whatever disks you have available. This is a cheap solution for testing iSCSI SAN scenarios.

rockygoldAuthor Commented:
Hi Brent,

Thanks for the valuable comments.I should clarify a few things. After this is being tested  in a lab setup, I want to consider this for a production setups with less than 100 users. I thought that the server hardware  we come across today, can handle such work loads of even 100 plus easily even at peak levels. My aim is to totally utilize all powered on servers while benefiting  from the HA features. At the same time I want to ensure that performance is not affected.
Are there any feasible ways to achieve these requirements in a production setup,or if not with high performance hardware?
Also one DC out of the two DCs will be installed in to one of  these two servers which participate in the cluster,however with no FSMO allocated to it but as a 2nd GCS. If performance is not an issue, which one of these clustered servers will be ideal as the 2nd DC in the cluster?
Also, as you said, openfiler is the best option just for the lab as I'm  comfortable with it.

Thank you
It is not recommended to put a DC on a cluster node. If you do have to do this however, you should put them on both nodes, not just one.
   "However, in a two-node cluster, if one node is a domain controller, the other node must also be a domain controller." Ref: http://support.microsoft.com/kb/281662

One thing that you could try is to put 2 VMs each on 2 systems (4 total VMs). 1 SQL and 1 Exchange VM per system and then cluster the VMs between the systems. That would eliminate the SQL/Exchange memory contention problems on the same system.
If you needed DCs, you could throw a DC VM on each system as well.

For what you are trying to do (server utilization), server consolidation by using VMs is probably the way to go.

Experts Exchange Solution brought to you by

Your issues matter to us.

Facing a tech roadblock? Get the help and guidance you need from experienced professionals who care. Ask your question anytime, anywhere, with no hassle.

Start your 7-day free trial
rockygoldAuthor Commented:
Brent, Thank you for the advice. I will  give it a go with  virtual/virtual first. Like you  initially mentioned, I will also test out physical/virtual get an understanding.thanks
It's more than this solution.Get answers and train to solve all your tech problems - anytime, anywhere.Try it for free Edge Out The Competitionfor your dream job with proven skills and certifications.Get started today Stand Outas the employee with proven skills.Start learning today for free Move Your Career Forwardwith certification training in the latest technologies.Start your trial today
Email Servers

From novice to tech pro — start learning today.

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.