WS-12-250DC 1.3.9 - Can't change voltage on the fly

[sirhc]

Well I like it when people speak their mind as that is how I am!

But the reason we put that safeguard in was people were sending units back RMA with burned out ports and of course they claimed they did absolutely nothing wrong.

However we do keep a sort of black box of information on the Flash of the last so many operations performed that we can access even if the board is fried so long as the flash chip is OK and in quite a few instances the user claimed they did not do this (change the POE from 24V to 48V or 24V to 24VH) and in fact that was in the black box log.

Therefore we came up with this safeguard and in fact the ONLY reason I can possibly think that you would ever want to change a POE option on a link that is working is this instance as UBNT AFX radios are the only radios I am aware of that can use 24V, 24VH, 48V, or 48VH.

Of course we would not be having this discussion on this event had UBNT not done what they have done before and change the POE voltage on a device.

But all you had to do was go port and turn POE off then change the POE option to whatever you wanted and the link would have came up.

But to be honest had you followed the manufacturer's recommendations you would have turned POE OFF on that port before unplugging the airMAX radio and plugging in the AFX radio then turned POE back on with the proper POE option but you obviously decided to plug and unplug live POE devices which is what we want people to not do. :(

[sakita]

And might I refer to my comment at the end of the aforementioned linked post...

I wonder if editing the config and then issuing a reboot command instead of trying to get it to apply it directly would work (and make a difference between success and lockup)? Or would it keep the PoE up during reboot and not change it? (which might explain the lockup as it would be something outside the intended operation of the switch).

Setting the PoE value up right the first time would also help - my point being that that is the first fat-fingering that happened - an indication of the human imperfection we all share (but, in this case, luckily not harmful as the equipment is working with the chosen setting). Given the demonstrated ability to make mistakes, having the switch prevent the higher likelihood damaging mistake by design is not a bad thing (and, yes, there are lawyers in my family).

Maybe a 20 minute drive this time but preventing a much bigger damaged equipment and swap out at another time. This reminds me of how people complain about traffic signals: they don't notice the several green lights they make it through but they sure notice that one red one. LIkewise we may not notice how many times the equipment saves our bacon but we sure notice that one time it annoys us by not letting us do something we really want to do.

Watching human behavior over the years (my own included) it is all too easy to bypass any popup warning boxes, captchas, or whatever - click, click, boom. I'm a very independent hands-on person that hates fighting with automagical technology but sometimes saving me from myself isn't a a bad thing.

[sirhc]

I wonder if editing the config and then issuing a reboot command instead of trying to get it to apply it directly would work (and make a difference between success and lockup)? Or would it keep the PoE up during reboot and not change it? (which might explain the lockup as it would be something outside the intended operation of the switch).

Messing with the config file is dangerous but what are you achieving by doing this?

When the switch reboots you will stop passing traffic for 30+/- seconds.

Now I guess if this is the only port/link that feeds the tower and if you change the POE to OFF you lose control of the tower which might be your hang up?

If that is the case you could modify the config file at your own risk and then reboot the switch yes, it should work but I would LAB the shit out of this procedure first on a bench.