Server 2008 R2 Multiple NICs, getting random RPC server unavailable errors
We have 3 servers that I am getting ready to create a failover cluster with, each with 8 NICs. I have configured the NICs according to their role, 4 for iSCSI that are plugged into the iSCSI switch, 1 on our backend network for management (Trust), 1 for Cluster communication, 1 for Live Migrations (it is going to be a HyperV cluster), and 1 for VM communication. During my initial configuration, I only have the Trust and 4 iSCSI interfaces active right now. My current IP config is: Host #1: Trust - 192.168.4.82/16 Def GW of 192.168.0.1 iSCSI#1 - 10.50.10.1/24 iSCSI#2 - 10.50.11.1/24 iSCSI#3 - 10.50.12.1/24 iSCSI#4 - 10.50.13.1/24 Host #2: Trust - 192.168.4.83/24 Def GW of 192.168.0.1 iSCSI#1 - 10.50.10.2/24 iSCSI#2 - 10.50.11.2/24 iSCSI#3 - 10.50.12.2/24 iSCSI#4 - 10.50.13.2/24 I currently have my binding order as: Trust, Cluster, iSCSI#1, iSCSI#2, iSCSI#3, iSCSI#4, Live Migration, Virtual Network. I am getting random errors when I query WMI from host #1 to host #2. Most of the time it works, but every so often I get RPC server unavailable, then it works again, then RPC errors. If I disable all NICs aside from the Trust interface (which is the only interface that registers with DNS), WMI queries work fine. I even get the RPC errors when I use the IP address of host 2, so it isn't DNS. It seems as if the WMI query is being sent out the wrong NIC interface, even though according to the binding order, it should be going out the Trust interface. Has anyone seen anything like this? I am relatively new to working with Windows 2008 R2, so it may be a config error on my side. Also, please don't advise me to post in the cluster or hyperv forums, as none of the servers are clustered yet, nor running HyperV. This issue is occurring during my initial server configs. Thank you Brian
January 25th, 2011 1:31pm

This topic is archived. No further replies will be accepted.

Other recent topics Other recent topics