Windows 2008 R2 cluster static or dynamically assigned IP addresses?
Posted on 2013-01-29
Windows 2008 clustering introduces the ability to use dynamically assigned IP addresses for the server nodes, whereas Windows 2003 required statically configured IP addresses. I can think of several reasons why this would be beneficial. For example, if I needed to change the primary DNS server address, I could change it in DHCP and avoid having to change it manually on potentially hundreds of cluster nodes throught our enterprise. In most cases, dynamic is better than static when managing a large enterprise.
Can anyone think of any disadvantages or problems with configuring cluster nodes and cluster names with dynamically assigned addresses? I still create reservations in DHCP for the resources so that they don't change and to avoid IP address conflicts. The only potential disadvantage I can think of would be if there was a failure of DHCP. We have a high availability solution for DHCP to help mediate this risk. Am I missing anything?