12-14-2018 06:51 AM
My US-16-150W tried to auto-upgrade to 4.0 on Monday and it appears to have killed the switch. It's constantly rebooting (going from blue led to flashing white to off, then repeat) and gives up after a few loops, with fans stuck on maximum. It will retry the rebooting loop if I powercycle it. Every once in awhile it's up for a minute or two and I'm able to login to fetch logs. The last thing I see is this:
daemon.info switch: BOXSERV: Error 17 occurred reading thermal sensor 3 data
daemon.info switch: BOXSERV: Error 17 occurred reading thermal sensor 4 data
daemon.info switch: BOXSERV: Error 17 occurred reading thermal sensor 5 data
daemon.info switch: UBNT_POE: PoE firmware check failed, reset the module. rc 0x000e
I've powercycled it a few more times and now I'm completely unable to ssh into it.
12-14-2018 08:47 AM
Did you have any power event that would have caused it to die when it was upgrading?
Just in case, I cancelled my auto-upgrade for tonight on my 16 150w. My WAPs all upgraded with no issue.
12-14-2018 11:53 AM
Sound like completely software issue.
Factory reset the switch to see if that helps, all configs will be pushed back to the switch once adopted by controller. If that does not correct the issue, I would try to ssh in long enough to kick off a firmware upgrade from the CLI to see if you can force re-flash the firmware.
12-14-2018 02:48 PM
I was able to ssh in for about 2 minutes and then the switch would reboot with the PoE firmware check message. It was never up long enough to be able to scp and apply a firmware update.
12-14-2018 08:10 PM
I just upgraded mine via the schedule and it had no issues. I'm not sure what happened to yours. I wish I could offer more help, but without having your switch in-hand, I won't be much help.
12-16-2018 05:19 PM
I unplugged all devices and was able to successfully upgrade it to 4.0. After plugging my devices in, plugging in any PoE device would cause the switch to go in to a bootloop.
Whatever went wrong appears to have killed the PoE module. Submitting an RMA now.
12-17-2018 10:32 AM
That's unfortunate. Hopefully they will get everything resolved! I've never seen a PoE Module cause complete failure of the flash. But I've dealt mostly with Cisco equipment.
12-17-2018 01:40 PM - edited 12-17-2018 01:41 PM
I Upgraded a Switch 16 POE-150W, Switch 8 POE-150W and Switch 8 POE-60W to 22.214.171.12453, thankfully none of them have experienced any issues whatsoever.
I did however read that version 4.0 firmware includes some improvements to the PoE module and to be patient in waiting for the switch to come back up as it could take a few minutes longer than a usual firmware update does.
2 weeks ago
I think my switch is doing same thing constant reboots once POE devices get applied to switch after a period of time
<14> Dec 31 19:00:38 UBNT UBNT_POE[Cnfgr_Thread ]: ubnt_poe_common.c(5114) 1 %% PoE firmware check failed, reset the module. rc 0x003f
a week ago
Jan 10 10:32:16 PoESwitch daemon.info switch: BOXSERV: Error 1 occurred reading thermal sensor 0 data
Jan 10 10:32:18 PoESwitch daemon.info switch: BOXSERV: Error 17 occurred reading thermal sensor 2 data
Jan 10 10:32:18 PoESwitch daemon.info switch: BOXSERV: Error 17 occurred reading thermal sensor 3 data
Jan 10 10:32:18 PoESwitch daemon.info switch: BOXSERV: Error 17 occurred reading thermal sensor 4 data
Jan 10 10:32:18 PoESwitch daemon.info switch: BOXSERV: Error 17 occurred reading thermal sensor 5 data
PoE restarts, knocking out our phones and AP-AC-L. Phones carry network to desktops, so it essentially shuts us down until the PoE restarts and phones restart.
a week ago
Thanks for this report. I will contact you via DM to gather more information from this switch.
Check out our ever-evolving Help Center for answers to many common questions!
FREE UBWA Student Guide-Great RF Primer!
FREE UBRSS Student Guide-Great Intro to Routing & Switching!
Having connectivity issues? See: UniFi Debugging Intermittent Connectivity Issues on your UAP
Learn How to Establish a Connection Using SSH Here