[Nut-upsdev] Tripplite USB 3003 problems

[Michal Hlavinka]

On Thursday 25 March 2010 02:25:38 Charles Lepple wrote:
> On Wed, Mar 24, 2010 at 11:35 AM, Wallis, Chase Civ USAF AFMC 519
> 
> SMXS/MXDEA <Chase.Wallis at hill.af.mil> wrote:
> > I have been having a few issues recently. I’m running Nut 2.4.3 on
> > Solaris 10 (10/09) using a TrippLite SMART700DVa UPS. Everything seems
> > to be working fine except for some random issues. During a power cycle
> > (UPS online to offline to online) NUT will occasionally get disconnected
> > from the UPS. A call to it using upsc results in “Error: Data stale”.
> >  And running dmesg has the following messages:
> > 
> > 
> > 
> >                 UPS ups at localhost on line power
> > 
> >                 UPS ups at localhost on battery
> > 
> >                 Data for UPS [ups] is stale – check driver
> > 
> >                 Poll UPS [ups at localhost] failed – Data stale
> > 
> >                 Last message repeated 2 times
> 
> After this, is the tripplite_usb process still running?
> 
> I don't know the default syslog configuration on Solaris anymore, but
> is there a more detailed system log than what "dmesg" prints? At the
> LOG_INFO level, the tripplite_usb driver prints out a message if it
> has given up on reconnecting to the UPS.
> 
> > This will only happen once in every 50 times or so. I am not sure where
> > to look for the problem. Could it be a problem in NUT? Or Solaris or the
> > libusb? Or could it be a problem with the UPS?
> 
> Even when the tripplite_usb driver actually works correctly, it is
> often attributable to a problem with the UPS :-)
> 
> All kidding aside, if this does not occur every time, it is going to
> be somewhat hard to track down. One possibility is to run the driver
> with some "-D" flags to generate debug output, but this will not work
> under upsdrvctl - you would simply start the driver by hand, and you
> will probably want to redirect the debug output to a log file.
> 
> > It has occurred on different
> > machines with different UPS’s, so I don’t think it is a hardware issue.
> > Has anyone else seen a problem like this? Or could point me in the right
> > direction to get this problem resolved?
> 
> I will admit that I do not have a tripplite_usb driver running
> continuously, and definitely not a protocol 3003 UPS, so we rely on
> user reports.

Hi,

one Fedora user has seen the same (or similar) problem with triplite_usb. 
After several power cycles ups driver goes crazy, probably hangs.

More info and log messages can be found here:

https://bugzilla.redhat.com/show_bug.cgi?id=575875

if you are interested in any info, let me know and I'll ask the reporter.

Cheers,
Michal