Re: Cluster Administrator and failover

From the terms and events you are describing, it sounds like you are using
both Network Load Balancing (NLB) and Failover Clustering on the same
servers. Those two solutions can't run together.

235305 Interoperability between MSCS and NLB
http://support.microsoft.com/default...b;EN-US;235305

--
Jeff Hughes, MCSE
Senior Support Escalation Engineer
Microsoft Enterprise Platforms Support (Server Core/Cluster)


"IT Admin" <> wrote in message
news:dc0d83f5-7158-457b-a07f-...
> Recently, we've rolled out a terminal server cluster with NLB. Seems
> to be working fine. So I though I would run some node failures to test
> what happens.
>
> This is where it get a bit weird.
>
> I currently have 2 nodes in the cluster. I can see the nlb ip address
> assigned to both servers as it should be. But when I shut down a node,
> things get a bit weird.
>
> First, I get event id's 1049
>
> Cluster IP Address resource 'Cluster IP Address' cannot be brought
> online because address 192.168.1.31 is already present on the network.
> Please check your network configuration.
>
> and event id 1069
>
> Cluster resource 'Cluster IP Address' in Resource Group 'Cluster
> Group' failed.
>
> But, when I look on the node that I didn't shut down, the IP address
> is assigned to it, the clussvc service is running, nlb manager shows
> the other node functioning correctly, and I can reach the server after
> a couple of minutes as the cluster responds to a node being shut down.
> If I bring the server back online, the cluster is once again fine.
>
> Right now, my best guess is that whatever cluster administrator reads
> when it checks on the cluster is messed up, because the cluster seems
> to be fine, I would like to resolve the problem though.
>
> Thanks