Re: [Apcupsd-users] Continual stream of battattach, battdetach, changeme and powerout events from BX1600MI

[Ted M. <te...@mi...>]

This kind of thing has come up before and the general recommendation has 
been to try the

UPS out on different physical hardware.  Not virtualized stuff which 
proves nothing.  Some people have

even plugged in a 4 port USB hub in between their PC and the UPS and 
fixed it that way.

USB is a very crappy standard and hardware and drivers are known to be 
buggy and incompatible in some

cases.  It's also quite possible that Powerchute's development team has 
already run into this problem and

added in logic to the software to ignore spurious junk on the USB bus 
because of these kinds of

incompatibilities.


Ted


On 2/20/2024 5:33 PM, Andy Smith wrote:
> Hi,
>
> On Thu, Feb 08, 2024 at 02:32:42PM +0100, Jernej Simončič wrote:
>> On Thursday, February 8, 2024, 01:20:37, Andy Smith wrote:
>>
>>> Have I configured something incorrectly here, or this model is known
>>> to not work with apcupsd, or it is faulty?
>> Something's wrong with your UPS, most likely the battery is bad,
>> but since you say the UPS is brand new, just get it replaced.
> I found some time to look more into this.
>
> Since the above date (and before), the Back-UPS BX1600MI has been
> supplying power to a few devices (~28% load) and aside from the
> spurious detach/re-attach and power fail/restore events, it has
> worked as expected. When I induce an event by unplugging mains, it
> does tell apcupsd this.
>
> If you recall, APC support would not discuss these spurious events
> with me unless I replicated the issues in their proprietary
> Windows-only Powerchute Serial Shutdown (PCSS) software.
>
> I instead contacted the vendor and told them I wanted to return it
> as faulty for a refund. They kindly offered to send an advance
> replacement out for me to then return the faulty one, so I took them
> up on that offer.
>
> In the mean time I found time to install a Windows VM and pass
> through USB to it, install PCSS there. Guess what? I can't replicate
> the problems with PCSS. It detects the UPS, shows all the stats,
> reacts to events as expected and doesn't generate any spurious
> events.
>
> You can probably see where this is going.
>
> The replacement Back-UPS BX1600MI arrived and it behaves the same
> way with apcupsd and nut: spurious events lasting less than 2
> seconds. I don't think two different instances of the same model are
> likely to be faulty in the same way.
>
> So, I'm wondering if maybe this particular model or group of models
> isn't faulty as such but does behave in a way that apcupsd doesn't
> like. Possibly these events really are being signalled by USB but
> PCSS is too slow to notice or something (?)
>
> Is there anyone currently using a Back-UPS BX1600MI or similar with
> apcupsd without these issues? Maybe I really have made some sort of
> configuration error, or there's something wrong with my USB setup in
> two different Debian Linux machines.
>
> Or is the BX1600MI using some different kind of protocol that
> apcupsd and nut don't understand? I am very willing to help debug
> this if anyone is willing.
>
> Can anyone help? As it stands I now have two UPSes that basically
> work but I can't usefully monitor and I can't prove they are faulty.
> :(
>
> Thanks,
> Andy
>
>
> _______________________________________________
> Apcupsd-users mailing list
> Apc...@li...
> https://lists.sourceforge.net/lists/listinfo/apcupsd-users
>