[f-nsp] configuring SX800 Ethernet interfaces in layer-2 mode

[Sujith]

Hello,

Thank you for your time and insight provided on the below issue.

One of our client is having Foundry SX800 as the core switch.  They got some
servers which they configured with Windows NLB to load balance across two
servers for some applications.

Both the servers are connected in the Foundry SX800 Ethernet ports.  This is
the common problem reported across the internet for this scenario (Windows
NLB and layer-3 switch.)

The Foundry layer-3 switch is not learning the cluster MAC address ( which
is configured in the servers) in both the Ethernet ports where it is
connected.

But the client is having a similar setup running on Cisco shop in another
location where they are not facing any issues.  The Ethernet ports where the
server's are configured in the switch running in layer-2 mode rather than
layer-3.  Another input got from my client is "the Cisco switch is not
seeing the Cluster MAC address" and so it is broadcasting to both the ports
when it is getting any request to the Cluster IP address.

I tried using Wireshark in the Foundry SX800 in the Ethernet port where i am
not seeing this MAC address but i am seeing only the Heart beat messages to
the Cluster MAC address from both the servers.  Apart from this messages, i
am not seeing any ARP request/reply through the Foundry Switch.

Does anyone come across these kind of issue and any work around for this
issue?

Please let me know if you have any queries where my above explanation is not
clear?

Thanks
-Sujith
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://puck.nether.net/pipermail/foundry-nsp/attachments/20100505/ec34150b/attachment.html>