Our organization's switches were placed onto the network sans configuration some time ago (long before I was hired). I decided to give them all basic configs with passwords and set up telnet access so we can manage them remotely. At one of our sites, a switch is exhibiting some bizarre behavior that's preventing me from enabling remote access. Here's the situation:
This site is on its own subnet. SW1 has an uplink to the router (a layer 3 switch), and links to two other switches, SW2 and SW3. All three layer 2 switches are Cisco 2950-24s. When I connected via console and configured them, both SW1 and SW2 had VLAN interfaces with IP addresses assigned via DHCP (which I then converted to reservations). SW3, however, did not receive one (I can't remember if it had a VLAN interface enabled and sadly don't have the config in front of me since it's not configured for remote access). Strangely, it *did* have an IP address---on one of the switchports. I didn't realize that was even possible for a switchport on a layer 2 switch to have an IP address, but it was right there in the running config. Even stranger, the port in question was the one attached to SW1 and it had the same address as SW1's VLAN interface(!).
So now if I try to add an IP address to the SW3 VLAN interface, the switch complains that it's on the same subnet as the rogue ethernet interface and refuses to add it. I haven't tried anything yet because I'm hesitant to make any changes during production hours---especially changes to a situation that contradicts so much of my CCNA training.
Does anyone:
A. remember seeing anything like this before?
B. have any idea what could cause such a scenario? or
C. know whether anything could reasonably be attempted while the switch is live?
