WS-6-Mini locking up in the cold (-15C)

[michaeln416]

Hi;

I have some WS-6-Mini's giving me a problem today; the ports seem to be locking up or cycling between off, 10/100, 1000 then back to 10/100 and off again.

We woke up to -15C temperatures and a number of support calls. In each case the issue was a WS-6-Mini powered by 48VH either from an AirFiber POE brick or from a WS-8-150-DC unit. Cable is Shireen FTP CAT5e with UBNT tough connectors. The chassis of the WS-6-Mini's are tied to ground, some have an ETH-SP (UBNT) and some do not. All are running firmware 1.4.8; Board Rev F; and all are encased in the Tycon plastic enclosure.

I have several units in the exact same configuration, in the exact same climate all doing fine. If this is temperature related, why is it only affecting a handful of devices and not all of them?

Anyone else run into this? Is there something that I can do to get these things to run a bit hot to fight off the cold?

Mike

[sirhc]

This is not cold issue in my opinion as these have been sold for 4 years and used in colder weather but I would swap them out with spares and see if problem stops.

If problem stops RMA for testing could be a defective part?

If problem persists then you know to look elsewhere.

[lligetfa]

Any chance it is the radios that are plugged into them that are flapping? I woke up to -33C this morning and nothing new in the logs.

How do the crimps do on the wiggle test?

[michaeln416]

Thanks for the replies. I went and investigated myself and in each case it was the ETH-SP that was the issue. All but one WS-6-Mini is working fine this morning at -25C. After I warm up I'll climb up and check out that one last device; my guess is that one is crimp related as there is no ETH-SP up there.

[MyThoughts]

We had a WS-12-250-AC reboot spontaneously last night around 11PM air temperature at the site was around -19C. The location in questions has 2 x WS-12-250-AC units only one rebooted. The unit that rebooted has about 135W of equipment while the other has about 85W on it.
At the time of reboot the board temps on the still operational unit was 36/45/33 for Board/CPU/PHY. The router at the tower location in the same equipment enclosure was reporting a Board temp of 11C and did not have any issues in the cold.

This reboot was at a different location the another WS-12-250-AC that started spontaneously rebooting every 12-20 hrs after upgrading to v1.4.8. I have been discussing that in another thread and at this time we are treating them as two different scenarios.

[sirhc]

Cold is not the cause of the reboots as these units have been sold and used in Canada and Alaska for 4 years in much colder environments. Although there could be a defective part sensitive to cold but I highly doubt it.

Please upgrade to v1.4.9 which was released last night and had an important fix for UBNT Discovery protocol that could cause issues.

If still rebooting after firmware upgrade then swap out switch with a spare switch and see if reboot continues, if reboots are still occurring then chances are there is something else afoot here.

If reboots stop after swapping with spare then RMA and we will figure it out.

[lligetfa]

Thanks for the replies. I went and investigated myself and in each case it was the ETH-SP that was the issue. All but one WS-6-Mini is working fine this morning at -25C. After I warm up I'll climb up and check out that one last device; my guess is that one is crimp related as there is no ETH-SP up there.

I am not surprised. The original run of the ETH-SP has a well known tolerance issue where even the slightest overcrimp results in a flaky connection. This is easily vetted by the wiggle test. Some of the same tolerance issue has shown up in a few AF5x radios.

[bchur83]

I am having a similar issue the last 2 weeks. The temps have averaged around 0F here in MN and I have one site that appears to have started random reboots the same day the temps dropped. Its a WS6 Mini in an outdoor enclosure so directly in the cold. The outside temp this morning is -13F and the entire switch is rebooting randomly at random intervals during the day.

I went to the site yesterday to check everything and it all looks normal. The Power Supply for the switch is plugged into a UPS and I tested pulling the power on the UPS and it held fine for a 5min test. I tried flapping the power to the UPS on and off fast but never dropped the output. I reseated all of the plugs/jack/cables and nothing looked out of the norm. It is powering a PowerBeam M5 and a Rocket M5 for a small POP, so no extreme loads or anything. The only thing I haven't tried is swapping the switch out yet.

I only found the issue when one of the customers off the tower reported some speed issues so when I started digging I saw that both radios had rebooted recently and then further looking it appeared they powered off/on at the same time and the AirControl 2 logs showed the history of it flapping since the cold hit here. I then looked back to last year during the winter season and noticed a trend that last winter it was doing the same thing over the winter time, I just hadn't noticed it going down since it would recover before AirControl would see the radios show offline.

I am going to try to swap switches today and see if the problem goes away or not. I will report back my findings.

[sirhc]

Our switches do not normally have any issues in the cold unless a part is defective (it happens)

Keep in mind there are known issues with some WS-6 and WS-8-150-AC units manufactured early last year and late 2016 where there was a defective CAP that caused reboots.

viewtopic.php?f=17&t=3308&p=21920&hilit=+known+issues#p21920
https://forum.netonix.com/viewtopic.php ... mon#p18176

If this is the case of your unit I apologize but we do warranty this issue for life.


However this is not known to be your issue.

Swap unit and if still occurs please post up more information such as:

Screen grabs of the following TABs
Status
Device/Status

Describe how you are powering it.

[lligetfa]

It is -37C here this morning. MINI is snug as a bug in a rug.