Fiber Backbone

[Guest]

Yes I do have a 3COM 4900 but this was added in hoping that it would fix the
problem. It did not fix the problem and its acting like a switch at this
time. Netbios must be running because I am not using wins and I have file and
print sharing enabled on the servers. About 45 computers are running XPPRO
and the others are running 98 both at standard installs. So probably they are
all in the election process. I am guessing they are. Most computers are at
100mb except for the workstations connected to switch [A]. They can not run
at 100 so I force the nic to run at 10. I did try running the program that
came with 3COM to map out your network and would show when things went down.
Running the program caused more downtime with the network. As I recall, the
utilization would jump up and redline. Then the network would go down. About
a year ago, I installed the latest firmware on all the switches and reset
them to the default setup. This did not help. I think I should be running
something that can sniff the network and see whats going on. What would be
the best least expensive program to run?

Looking at the 3COM port stats, the worst ports are the fiber that one has
390 CRC errors. And WOW, switch [C] ports 20 has 231137 undersize and 241548
Collisions and 22 has 17218 undersize 19973 collisions. no CRC errors. A
quick browse of the other switches looks fine. Cant think no more getting too
late.

 

[Herb Martin]

Yes I do have a 3COM 4900 but this was added in hoping that it would fix the
problem. It did not fix the problem and its acting like a switch at this
time.

Switch comes in (at least) two types plus hybrids
so saying "like a switch" is pretty vague:

1) Bridge layer-2 switch
(propagates all broadcasts by default)

2) Router layer-3 switch
(doesn't propagate broadcasts by default)

Netbios must be running because I am not using wins and I have file and
print sharing enabled on the servers.

File sharing can work without NetBIOS -- browsing
doesn't.

About 45 computers are running XPPRO
and the others are running 98 both at standard installs. So probably they are
all in the election process. I am guessing they are. Most computers are at
100mb except for the workstations connected to switch [A]. They can not run
at 100 so I force the nic to run at 10. I did try running the program that
came with 3COM to map out your network and would show when things went

down.

You likely have DHCP failing due to a switch
problem that is failing to propagate the broadcast
(Discovery) OR which is acting as a router and
(properly) blocking them.

Running the program caused more downtime with the network. As I recall, the
utilization would jump up and redline. Then the network would go down. About
a year ago, I installed the latest firmware on all the switches and reset
them to the default setup. This did not help. I think I should be running
something that can sniff the network and see whats going on. What would be
the best least expensive program to run?

I would start with the (FREE) NetMon on the
DHCP server and see if any broadcasts or renewal
requests for DHCP are reaching it from the affected
DHCP clients.

(Reboot or use ipconfig /renew to get some activity.)

Looking at the 3COM port stats, the worst ports are the fiber that one has
390 CRC errors. And WOW, switch [C] ports 20 has 231137 undersize and 241548
Collisions and 22 has 17218 undersize 19973 collisions. no CRC errors. A
quick browse of the other switches looks fine. Cant think no more getting too
late.

 

[Guest]

Thanks for all of your time. I have learned enough to get me started and see
what I can fix my network. On my next reply, I will begin a new subjet Fiber
Backbone 2

Thanks and both of you have a great weekend.

Yes I do have a 3COM 4900 but this was added in hoping that it would fix the
problem. It did not fix the problem and its acting like a switch at this
time.

Switch comes in (at least) two types plus hybrids
so saying "like a switch" is pretty vague:

1) Bridge layer-2 switch
(propagates all broadcasts by default)

2) Router layer-3 switch
(doesn't propagate broadcasts by default)

Netbios must be running because I am not using wins and I have file and
print sharing enabled on the servers.

File sharing can work without NetBIOS -- browsing
doesn't.

About 45 computers are running XPPRO
and the others are running 98 both at standard installs. So probably they are
all in the election process. I am guessing they are. Most computers are at
100mb except for the workstations connected to switch [A]. They can not run
at 100 so I force the nic to run at 10. I did try running the program that
came with 3COM to map out your network and would show when things went

down.

You likely have DHCP failing due to a switch
problem that is failing to propagate the broadcast
(Discovery) OR which is acting as a router and
(properly) blocking them.

Running the program caused more downtime with the network. As I recall, the
utilization would jump up and redline. Then the network would go down. About
a year ago, I installed the latest firmware on all the switches and reset
them to the default setup. This did not help. I think I should be running
something that can sniff the network and see whats going on. What would be
the best least expensive program to run?

I would start with the (FREE) NetMon on the
DHCP server and see if any broadcasts or renewal
requests for DHCP are reaching it from the affected
DHCP clients.

(Reboot or use ipconfig /renew to get some activity.)

Looking at the 3COM port stats, the worst ports are the fiber that one has
390 CRC errors. And WOW, switch [C] ports 20 has 231137 undersize and 241548
Collisions and 22 has 17218 undersize 19973 collisions. no CRC errors. A
quick browse of the other switches looks fine. Cant think no more getting too
late.