[Nut-upsuser] nut upset after Oops

Arjen de Korte arjen at de-korte.org
Tue Mar 6 17:11:59 CET 2007


>> Please start by providing some information. What NUT version are you
>> using, what driver?
> The FC6 binary for 2.0.3-2.1.
> I use the Mustek PowerMust driver with a Sweex UPS.

Before trying anything else, upgrade to nut-2.0.5. There have been many
bugfixes since nut-2.0.3, including problems with bogus staleness warnings
and connection losses. This includes switching to the 'megatec' driver.

>> What messages do you see? How frequently?
>
> Mar  6 16:21:12 epia upsd[21151]: Data for UPS [myups] is stale - check
> driver
> Mar  6 16:21:14 epia upsmon[21155]: Poll UPS [myups at localhost] failed -
> Data stale
> Mar  6 16:21:14 epia upsmon[21155]: Communications with UPS
> myups at localhost lost
> Mar  6 16:21:14 epia wall[22834]: wall: user nut broadcasted 1 lines (46
> chars)
> Mar  6 16:21:15 epia upsd[21151]: UPS [myups] data is no longer stale
> Mar  6 16:21:19 epia upsmon[21155]: Communications with UPS
> myups at localhost established
> Mar  6 16:21:19 epia wall[22852]: wall: user nut broadcasted 1 lines (53
> chars)
> Mar  6 16:21:24 epia upsd[21151]: Data for UPS [myups] is stale - check
> driver
> Mar  6 16:21:24 epia upsmon[21155]: Poll UPS [myups at localhost] failed -
> Data stale
> Mar  6 16:21:24 epia upsmon[21155]: Communications with UPS
> myups at localhost lost
> Mar  6 16:21:24 epia wall[22886]: wall: user nut broadcasted 1 lines (46
> chars)
> Mar  6 16:21:26 epia upsd[21151]: UPS [myups] data is no longer stale
> Mar  6 16:21:29 epia upsmon[21155]: Communications with UPS
> myups at localhost established
> Mar  6 16:21:29 epia wall[22889]: wall: user nut broadcasted 1 lines (53
> chars)
> (etc)

This is most likely a server (upsd) problem and probably fixed in
nut-2.0.5 (the logic for determining if a driver is stale was flawed).

>> Does
>> unplugging and replugging the UPS help?
> Didn't try that.

This might work, since it will probably make the driver signal a state
change (which the server could have missed). This is a major hack though
and really should be solved by upgrading to a more recent version of NUT.

>> It could be as simple as you didn't set pollinterval correctly in your
>> ups.conf, or MAXAGE in upsd.conf, or DEADTIME or POLLFREQ in
>> upsmon.conf (see ups.conf(5), upsd.conf(5), upsmon.conf(5)). Or it
>> could be something more complicated.
>
> I already adjusted those some time ago after suffering messages when USB
> activity occurred.

To which values?

> Of course the oops may be a factor here but that UPS status does not
> change, really.

As said, the problem is probably in the connection between driver and
server, so the state of the driver probably doesn't change.

Best regards, Arjen
-- 
Eindhoven - The Netherlands
Key fingerprint - 66 4E 03 2C 9D B5 CB 9B  7A FE 7E C1 EE 88 BC 57




More information about the Nut-upsuser mailing list