Windows Server 2008 Data Centre - Failover cluster : un-cluster

Question

We have a pair of powerful servers configured for fail-over, using windows servers built in fail-over clustering and hyper-V sharing san storage.

The configuration is slightly jagged as there is a Windows service running on one of these hardware nodes outside of the cluster. We then have 3 virtual machines (each windows 2008 r2 server) that are managed within the cluster, a web server, a sql database, and an application server.

All of our processing requirement fits comfortably in one hardware node, and I have recently had a hard time defending the hardware redundancy. So fail-over is now considered an expensive luxury, which has never been used.

I need to split apart (un-cluster) the servers, and run on a single hardware node (for now). This is a live environment.

What gotchas will I face?

What do I need to be aware of if anything?

Thanks.

Answer 1

All you need to do is move the three virtual machines to the box that will be running the system, then evict the other node.

As for justifying having the redundancy, that's easy. Put everything on one node and power off the other node. Then tell your boss that the production server needs an emergency hotfix and that production will be down for 10 minutes while the server reboots and the VMs all go down. Odds are that the redundant server will be justified pretty quickly.

The fact that the clustering failover doesn't appear to have ever been used is a good thing. I'm guessing that it has been used, probably monthly when the new patches some out or when ever you install Windows updates. It isn't costing you a lot of money to keep the box up and running every month (basically the power and cooling are your monthly costs). This allows you to patch the hosts easily without needing to wait until the middle of the night to do it saving staff (you) many headaches and late nights which can be easily avoided.