Looking for help on this as I'm completely stumped.
An industrial customer has 4 Ubiquiti Edgeswitch-48-500w's running a camera network. One switch keeps 'freezing' about once every 7-14 days. When this happens neither the switch nor any of the hosts on the switch are reachable. Average throughput on this switch is 200-300mbps, it is daisy-chained (number two position) off one other switch (other two switches are on separate network). I RMA'd the bad switch and got a new one. The new switch is having the exact same problem as the original switch. I have to hard-power cycle the switch to get it back online.
What could be causing this with two different switches? After a reset all the hosts come up and run fine, but would a bad termination / short somewhere cause this? Any recommendations on how to troubleshoot?
Unfortunately, the onboard logs aren't offering any help. I just dialed them up to 'debug' to see what I can find.
One catch: on Oct 24th (last freeze event) I installed a network monitoring power outlet to cycle the switch's power when the outlet loses contact with the router (via ping). When the switch froze today, the outlet did not power cycle the switch and I was surprised to find I could still access the outlet's web GUI and had to manually power cycle the outlet. The power outlet's NIC is connected to the bad switch. Now I'm even more humbled and confused.
I can't image this would matter, but ports 1-38 are 100mbps poe connections to cameras drawing about 3.8 watts average / 144 watts total. Port 47 is used by the power outlet, port 48 is the trunk / daisy chain to switch 1. No VLANs / flat network. Switch #1 is also connected to cameras, as well as the server and router, throughput is ~350mbps.
e.g.: Bad switch -> switch 1 -> router
No problems at all with switch #1.
All help / insight / troubleshooting ideas are welcome!
No comments:
Post a Comment