Cable test didn't show issue, but turning on PoE made smoke.

[slimey]

Following damage to one of our PoPs over the weekend from lightning which seemed to kill the Netonix WS-12-250-DC we were using, I swapped it out with a new one, and was cautiously running a cable test before enabling PoE (24v) on a port. First two ports I tested, Cable Diagnostics showed all 4 pairs as "open", with a length of each pair as 9m (about the height of the pole). Turning on PoE seemed to work, but no link light, so radio diagnosed as dead.

Tried this on a third port (which was actually port 1), Cable Diagnostic showed same - all pairs as open, and correct length - so I enabled 24v PoE. Crackling noise came from the switch, and some smoke escaped, so I very quickly disconnected the RJ45 from the socket. Switch otherwise appeared to continue intact, but I assume I've fried the PoE circuit on this port.

I see from the logs that it did its own cable test before turning on the PoE:


Dec 31 19:22:47 switch[1388]: PoE enabled on port 1, PoE Smart is starting cable check
Dec 31 19:22:52 switch[1388]: Port 1 cable check results: Open, Open, Open, Open
Dec 31 19:22:52 switch[1388]: Port 1 passed PoE Smart cable check, turning on power

Clearly the cable check is missing something here, and didn't adequately protect the port. Does this count as a hardware bug?

Radio was a Ubiquiti Rocket 5AC Prism. Replaced it with a new radio (but same cabling), and it's now working fine (but on a different port on the switch).

Simon

[sirhc]

You should never see all OPEN on any port with a radio or other ethernet device.

The only way you see OPEN is there is nothing on the other end if the cable.

What you would see is OK not OPEN.