I am managing a small Novell NetWare environment that consists of: (2) Netware 6.5 SP6 boxes, (7) Windows XP SP3 workstations, and (2) network printers. These devices are all connected to a basic, unmanaged, 10/100/1000 Mbps switch (Netgear JGS524).

The Issue:
Unexpectedly, all devices on the Network will randomly lose connectivity at the same time (no pings go thru to/from any device), and the lights on the front of the switch "freeze" in their current state (for example, if I pull a network cable out of a switch port, it will remain illuminated in its last state, as if the cable were still plugged in). The only way to restore connectivity at this point is to reset the switch by cycling the power on/off.

Troubleshooting History:
We've been experiencing this problem for quite some time, and at first I assumed the switch was bad, so I replaced it with a different model only to find that the issue still occurs (the switch listed above is the current model we're using).

I'm having trouble diagnosing this problem, as I can't seem to correlate what's happening on the network to initiate the switch lock-up. Sometimes it will happen twice in one day, and sometimes it will go 3 or 4 days with no problems. It would be hard to leave a packet scan running for 3 days and then try to determine at what time the lockup occurred if Im not continually observing the switch 24/7.

Anyway, any thoughts on a good approach to narrowing down what's causing this? Or any obvious culprits that might exist in a Netware environment that could be causing a broadcast storm?