NLB Networking Issue (2 Node TS Cluster)

I have spent WAY too long trying to figure this out...

I have two terminal servers configured in Unicast mode with separate
NIC's. The cluster nics are configured to 10.25.1.35 using the NLB
Manager. The other NIC's are on a separate subnet 10.26.1.11 and
10.26.1.12.

Initially, everything was plugged into the same switch (Dell
PowerConnect 5324). I know there are some MAC issues with switches, so
I've also tried pluggin all 4 cables into a 3Com hub and uplinking it
to one port on the Dell switch, and that makes no difference at all.
I've even set the MaskSourceMac reg key to 1 as well.

My Problems:

1. Server 1 (x.11) can access the internet and ping other servers on
the 10.25.1.x subnet (thru the cluster NIC) and also access Server 2
via the 10.26.1.x subnet, however when using the NLB Manager to view
the cluster it fails to connect to (or properly resolve)
Server2.Domain.Com.
2. Server 2 (x.12) can only access Server 1 via it's IP address thru
the 10.26.1.x subnet. Trying to ping any server names on either
network does not work.
3. More interesingly, if I pull or disable the cluster NIC on Server
1, Server 2 will then take over and I can access the internet and ping
anything I want!
4. A side note, pinging other servers on the network with the "ping
x.x.x.x -t" command is steady and normal on Server 1 however on Server
2 the timing fluctuates??? I've tried multiple NIC's to rule out
hardware.
5. I'm guessing my problem lies with the switch being confused with
the MAC addresses somehow. I've managed to connect to the Dell switch
and attempted to add and static ARP entry, but I am not sure if this is
the right thing to do. Perhaps a VLAN is the way to go...?


Thanks!

have you tried running the NLB cluster in multicast mode? (I don have a lot
of experience in this area but I know that this solves some issues like
switch flooding).

--
Regards,
Andrei Ungureanu
www.eventid.net
Test our new EventReader!
http://www.altairtech.ca/eventreader...lt2.asp?ref=au

<(E-Mail Removed)> wrote in message
news:(E-Mail Removed) ups.com...
>I have spent WAY too long trying to figure this out...
>
> I have two terminal servers configured in Unicast mode with separate
> NIC's. The cluster nics are configured to 10.25.1.35 using the NLB
> Manager. The other NIC's are on a separate subnet 10.26.1.11 and
> 10.26.1.12.
>
> Initially, everything was plugged into the same switch (Dell
> PowerConnect 5324). I know there are some MAC issues with switches, so
> I've also tried pluggin all 4 cables into a 3Com hub and uplinking it
> to one port on the Dell switch, and that makes no difference at all.
> I've even set the MaskSourceMac reg key to 1 as well.
>
> My Problems:
>
> 1. Server 1 (x.11) can access the internet and ping other servers on
> the 10.25.1.x subnet (thru the cluster NIC) and also access Server 2
> via the 10.26.1.x subnet, however when using the NLB Manager to view
> the cluster it fails to connect to (or properly resolve)
> Server2.Domain.Com.
> 2. Server 2 (x.12) can only access Server 1 via it's IP address thru
> the 10.26.1.x subnet. Trying to ping any server names on either
> network does not work.
> 3. More interesingly, if I pull or disable the cluster NIC on Server
> 1, Server 2 will then take over and I can access the internet and ping
> anything I want!
> 4. A side note, pinging other servers on the network with the "ping
> x.x.x.x -t" command is steady and normal on Server 1 however on Server
> 2 the timing fluctuates??? I've tried multiple NIC's to rule out
> hardware.
> 5. I'm guessing my problem lies with the switch being confused with
> the MAC addresses somehow. I've managed to connect to the Dell switch
> and attempted to add and static ARP entry, but I am not sure if this is
> the right thing to do. Perhaps a VLAN is the way to go...?
>
>
> Thanks!
>