Networks not chatting

Try the easy things first. Start by power cycling the switch. Next thought, is there a potential single point of failure, such as a single cable going connecting the router to the switch on 192.168.100? Replace it with new or known good. If that doesn't work, plug it into a different port on the switch and try again.

When you don't know the cause, go from easier to harder.. Let us know how this turns out!

not enough information to even quess.

What internet service do you have? how does it come in? what type of routers are used to segment [subnet] the networks?

Here is an official report I drummed up. Note that IP addresses and such have been changed for security reasons, which is why I leave redundant notes explaining what the same IP addresses do.

We have 2 seperate networks running through our campus. One is for staff/faculty and the other is for students/public computers. The staff can access e-mail/internet, and the students cannot. This problem happened over the weekend, nobody was in to change any settings/cable runs, but there was a power outtage Friday night. Here is the raw data of the network...

Edenserver01 (our main file and exchange server) has 2 NIC cards, the "Faculty" side is 10.1.1.210. The "Student" side is 192.168.100.2.
Eden-webserver (our web server) has 2 NIC cards, 1 side is 10.1.1.204. The other is 192.168.100.5.
We are unable to access the blackboard server because of a password that nobody knows, but we do know it is 65.16.18.179.
Accessserver is 10.1.1.52

Every NIC with an IP starting with 192 has a Gateway of 192.168.100.1. NIC's that start with 10 have a gateway of 10.1.1.201. All subnet masks are 255.255.255.0.

All of the wiring closet machines are plugged into switches on ports without VLAN's. Except the unnaccessable blackboard server, it's on a port with a VLAN configured. All of the wiring closet machines are on 1 switch, except the blackboard server. All of the other wires in the switches are unlabled and their sources are unknown.

Edenserver01 and Eden-webserver are the only 2 PC's on campus that can ping 192.168.100.1. They are also the only 2 machines that have both a 10.1.1.x and 192.168.100.x NICS. Every machine can ping 10.1.1.201, except the computers on the student/public network.

We switched some cables around, brought the 192.168.100.1 cable (gateway) and the 192.168.100.2 cable (Student from Edenserver01) down to the bottom switch, also on unassigned ports. The internet went back up for the students, but outgoing mail (such as to yaho

yahoo.com, eden.edu still worked.) stopped going out, and instead just queued up every message in SMTP. Incoming mail progressed as usual.

192.168.100.1 (the student network) leaves through 65.16.81.178. 10.1.1.201 (the faculty network) leaves through 65.16.81.178. Nothing can ping 65.16.81.178 (the port to the outside on the student side.) Only Edenserver01 can ping 65.16.81.174.

Event Viewer tells me that their are duplicate IP addresses on the network, one of which belonged to a Dennis Ammann at 10.1.1.82. He is no longer with us, his office has a completely different computer/IP address, and yet it still says he is currently on the network.

I went to where Dennis's computer used to be. It was occupied by a woman who had a laptop instead of a desktop. The NIC card was set to "detect settings", and she was assigned an IP address of 192.168.100.22, with a gateway of 192.168.100.1. She had connectivity, and is the only person on the 192 network with connectivity (that I have found.) The woman in the office next to her had a static DNS that led to an outside line, but a "detected" IP of 192.168.100.63. She could not connect. I set her up with a static local DNS, unsuccessfully. I set her up to detect all settings, which also failed.

Our networks are connected by 2 T1 lines. They run into a Cisco Catalyst 2600 router.

I messed up while changing the IP's around and used the same one twice... It should read,

10.1.1.201 (the faculty network) leaves through 65.16.81.174.

Interesting configuration because the Edenserver01 is multihomed, yet it allows both the faulty and students to access it but on seperate subnets. Further, you've got managed switches, patch panels and an access router.

If students require internet access, there has to be a gateway out of their 192.168.100/24 subnet. If you specify the 192.168.100.1 as the gateway, you have to find out where that is. That has to be a NIC somewhere [or a configured uplink port] which gets connected to the Access router [your 2600] which is connected to the internet. Could be simply a uplink crossover cable failed. could be that with the power outage when everything came back online, something wasn't recognized as being there [happens to us on a W2003 server and a catalyst 1900. DHCP doesn't see the catalyst because the server comes up faster than the switch can go through its diagnostic and get the ports active. Therefore the server sees the NIC attached to it as unconnected and DHCP cant initialize].

Just some thoughts.

If you are running 2 networks there is probably a router somewhere to tie the two networks together. Try some pathping commands from each network and see were the responses die.