dhcp 3com managed switch problem

[Guest]

Hi all,
I just got a 2250 plus 3com switch that can be managed. I assigned a static
ip address 192.168.1.49.
I am having problems with it not assigning ip address using dhcp.
My setup is:
2000 server running dhcp with the following scope:
ips 1-50 (are used for static)
included 192.168.1.50 thru 192.168.240
excluded 192.168.1.50 thru 192.168.150

2003 server running dhcp with the following scope:
included 192.168.1.50 thru 192.168.1.240
excluded 192.168.1.151 thru 192.168.1.240
The purpose of two dhcp servers was to have a failover if one server was down.

the 2000 server shows to have 100% ips used when in fact it only show about
10 assigned from the 2000 server.
DHCP seemed to be working fine until I got the 2250 plus managed switch.
Should the ip of the switch be within a scope for some reason or can it be
outside the scope just like my servers are?
What is happening is when I plug in a pc to a port on the 2250 switch it
does not find a 192 address and reverts the ip of the pc to a 169 number and
it will not connect ot the network.
If I assign a static ip to that pc and connect then it works then if I
change back over to auto dhcp then it will start working.
The switch seems to be working fine so I think it might have to do with my
dhcp scope. I have done three pcs and they all connect to my 2003 server.
Any suggestions?
Sher

 

[Kurt]

Hi all,
I just got a 2250 plus 3com switch that can be managed. I assigned a static
ip address 192.168.1.49.
I am having problems with it not assigning ip address using dhcp.
My setup is:
2000 server running dhcp with the following scope:
ips 1-50 (are used for static)
included 192.168.1.50 thru 192.168.240
excluded 192.168.1.50 thru 192.168.150

2003 server running dhcp with the following scope:
included 192.168.1.50 thru 192.168.1.240
excluded 192.168.1.151 thru 192.168.1.240
The purpose of two dhcp servers was to have a failover if one server was down.

the 2000 server shows to have 100% ips used when in fact it only show about
10 assigned from the 2000 server.
DHCP seemed to be working fine until I got the 2250 plus managed switch.
Should the ip of the switch be within a scope for some reason or can it be
outside the scope just like my servers are?
What is happening is when I plug in a pc to a port on the 2250 switch it
does not find a 192 address and reverts the ip of the pc to a 169 number and
it will not connect ot the network.
If I assign a static ip to that pc and connect then it works then if I
change back over to auto dhcp then it will start working.
The switch seems to be working fine so I think it might have to do with my
dhcp scope. I have done three pcs and they all connect to my 2003 server.
Any suggestions?
Sher

"the 2000 server shows to have 100% ips used when in fact it only show
about 10 assigned from the 2000 server"

Shows to have 100% IPs used where? All of the leases are used? Or do you
mean it acts like they are all used up because your other network
clients can't get a DHCP address?

1) You need to isolate the problem to a switch problem or a DHCP
problem. Put the old switch back and see if the problem goes away. Look
at your event logs on the DHCP server, and look in the DHCP snap-in to
see if you are indeed exhausting your DHCP addresses. If all is well
with an unmanaged switch, go to step 2.

2) Are your computers plugged directly into the new 3-Com? Try enabling
spanning-tree portfast (That's Cisco terminology, but 3Com will have an
equivelant). You wouldn't expect a port to take THAT long to begin
forwarding traffic, but portfast mode will ensure your DHCP requests are
not timing out waiting for the switchport to wake up. If you're not
using redundant links from that switch, you can disable spanning-tree
altogether.

Post back your results.

....kurt