Ok first to start out, I think everything is configured right and the results
i'm recieving are expected.
I have 4 nodes in my NLB cluster each with two network cards. The NLB NIC
has a cluster IP and a second unique IP. The second NIC has just a unique IP.
All 9 IP addresses are on the same subnet and share the same Gateway address.
All register to the same DNS server and are located on the same switch.
From one node on the NLB cluster I can not ping another node. In NLB manager
I can see some or none of the other nodes in the NLB cluster.
When I do ping tests from other machines that are not in the NLB cluster,
everything pings back fine. When I connect NLB Manager to the cluster from a
machine that is not part of the NLB cluster, all nodes are present with Green
lights.
Everything is set to Unicast and I think I prefer it that way. From what I
have been reading, multicast will eliminate this issue, but opens a whole new
can of worms.
Is this typical for this configuration? Can I leave it this way? Everything
seems to be working fine. I ran the command line utility and everything
checks out. I’m looking for additional opinions to confirm that this is
typical and that I can leave the configuration the way it is.
Thanks,
Amit
|