I have a 3 node Netware 6x Cluster. Each Dell server has dual Nics which are configured as a team - 1 IP load balanced using Novell's built in functionality. This has been in place for 4 years. Last year a network upgrade took place and 3com gear is in place. We have had reports of "slowness" from various points in our network-not for accessing these servers specifically. The 3Com guys are pointing the finger at the Novell Cluster saying it is misconfigured. Below is what they are saying - not being a network guru I'm not sure how to go about disputing this. Any feedback would be appreciated. Hopefully I'm posting in the right forum.

Consulting with my senior engineer we discussed the "bleeding" of
unicast traffic. We are concerned that the MAC tables and ARP Cache
tables are being re-freshed. While this re-fresh is taking place it is
possible for unicast bleeding to take place. We want to eliminate possible causes of the refresh. We want to disconnect one connection to each of the dual home servers causing the duplicate IP addresses in core F1. This problem must be addressed any case. The servers have team or dual home NIC cards for redundancy. They have 2 different IP addresses reporting to same MAC address. This can cause SPT to refresh MAC and ARP tables. Your server IT personnel should contact manufacture of NICs for proper configuration.