Hello, and many thanks in advance for any advice you might have.

I adminster small, two node, deparmental server cluster on a large
university LAN. Our servers are running Netware 6.0 SP5. I am getting the
following error message:

7-17-2006 11:27:20 am: TCPIP-6.10-78
Severity = 1
This Server and the system having hardware address 00-02-B3-ED-CF-6B have
conflict for IP address 128.59.xxx.xxx.

where the hardware address listed is the server's own NIC. The IP address
hand entered into the server's config file.

This has happened intermittently for several months and I asked for help in
March on this forum. I was advised that it probably indicates that the
Spanning Tree Protocol was disabled on the switch. Then I stopped seeing the
error message.

Before I go any farther, is this a potentially serious issue? If it is just
cosmetic I'll live with it. (The only problem that seems it might be related
is sometimes a PC may take overly long to login on the cluster,
occassionally to the point of timing out and having to restart.)

When I noticed the error was back yesterday, I contacted our network
infrastructure group and asked them to check the switch. They confirmed that
STP is enabled on the switch and the two ports that our machines are
connected to are in a Spanning Tree Protocol forwarding state.

It doesn't seem to matter which server is the primary at the time. Our
preferred server went down last week due to a CPU hog and the other took
over without a flaw, but the IP conflict error popped up within a few
minutes on that server.

I also looked at TID10016507 but it says this frequently an issue with the
switch being set to Full Duplex and the server to Half Duplex. Our servers'
NICs are set to Full Duplex.