OK this will be short because it’s Sunday after all, and despite that I have a crushing deadline. However, I wanted to share a problem with clustering VMs that I’m sure many others will also encounter & for which I found a solution.
I’ve created many clusters in the past, and needed to create one for a lab environment. I used FreeBSD to create NAS storage. Though StarWind would have worked fine and i’ve used it many times in the past, i decided to do something different and more elaborate. Plus it’s just fun to learn something new.
Anyway, so then I pointed two W2K8 R2 servers to the NAS storage using iSCSI, installed Failover Clustering feature, and tried running validation tests. (I didn’t care too much if warnings appeared in validation – lab environments without “true” server tier hardware sometimes throw warnings.) However, when trying to create the cluster, i ran into this error:
“The operation has failed. An error was encountered while loading the network topology. Error Code: 0x80070005”
Wow, well that stumped me. I had the two NICs on both servers properly configured for two separate networks & with the heartbeat network second in the binding order. What could be wrong with the network topology? I chased around various network troubleshooting scenarios for an embarassing length of time, did a Google search where I found someone else with the same problem but no real solution other than reverting to W2K3, (http://capitalhead.com/articles/failover-cluster-validation-error-80070005-on-windows-server-2008-r2-x64.aspx) and lost a good night’s sleep mulling this over.
Have you ever known that you’ve spent way too much time on troubleshooting something and long ago passed the point of diminishing returns, but because of a curious and persistent nature, doggedly pursue a solution? That was me. I was frustrated and embarassed that i should have been able to get this figured out by now.
OK so i found the answer. Many times to save disk space, we’ll create a W2K8 VM and if we need an additional server, create a second one based on the first. The second one is a clone. (Working in VMWare workstation here.) Anyway, i’ve been able to make this work just fine for many server demonstations & testing. Turns out, the networking error that i was seeing had nothing to do with networking as the error message would have me believe. Because i was clustering two cloned machines, they both had the same SID. I ran c:windowssystem32sysprep.exe /generalize on one of the machines to give it a unique SID. Thereafter, clustering worked like a charm!
Turns out I would have found the solution earlier, except that my Google search was actually TOO SPECIFIC and returned only one hit in the link shown earlier. Later, using a more generic search, I found this http://networkerslog.blogspot.com/2010/03/when-i-run-validation-of-my-test-hyper.html?showComment=1303081576346#c5390058544218802676 which pointed me in the right direction (although it failed for many other people…?). Thanks samyee of Singapore who posted this.
Finally, here are a couple of great links with instructions in broken (but understandable) English that configure your storage using FreeBSD and also cluster with W2K8 R2