Charles Lepple
2012-Jul-21 17:23 UTC
[Nut-upsuser] Fwd: Re Powercom Vanguard driver for nut
resending to the list - got an error the first time. From: Charles Lepple <clepple at gmail.com> Subject: Re: [Nut-upsuser] Re Powercom Vanguard driver for nut Date: July 21, 2012 1:21:22 PM EDT To: Gordon Longland <gordon at compsos.com.au> Cc: nut-upsuser at lists.alioth.debian.org <nut-upsuser at lists.alioth.debian.org> On Jul 21, 2012, at 7:52 AM, Gordon Longland wrote:> Network UPS Tools - UPS driver controller 2.2.2 > Starting UPS: UPS > exec: /sbin/usbhid-ups -a UPS -u nut > Network UPS Tools: 0.29 USB communication driver - core 0.33 (2.2.2)Support for this PowerCOM device appears to have been added in NUT 2.6.3 (according to the NEWS file).> Jul 21 21:32:29 server upsd[3760]: Can't connect to UPS [UPS] (usbhid-ups-UPS): No such file or directory > Jul 21 21:32:30 server upsmon[3764]: Poll UPS [UPS at localhost] failed - Driver not connected > Jul 21 21:34:35 server upsmon[3764]: UPS UPS at localhost is unavailable > > Is there something simple we are missing? Why does it add the UPS stanza to the driver (usbhid-ups-UPS) name and then fail?The name in square brackets is the name for that specific UPS instance in ups.conf, and the driver name (usbhid-ups) is prepended to form the name of the socket which the driver and 'upsd' use to communicate. If the driver hasn't started successfully, upsd can't find that socket.> Is there a patch for the usbhid-ups driver to recognise the Vanguard series?It might be possible to patch in support, but you can probably find a RPM build of a newer version of NUT, or you can try building it from source. Email us back if you have questions about either method. -- Charles Lepple clepple at gmail
[please keep the list CC'd, most likely with reply-all] On Jul 21, 2012, at 6:46 PM, Gordon Longland wrote:> Thank you for the reply Charles > > https://bugzilla.redhat.com/show_bug.cgi?format=multiple&id=806567 > > The above bug reflects the issues I had with trying to compile nut on RHEL5 i686 system. Has anyone managed to get nut 2.6.3 working on RHEL5?I don't currently use any RPM-based systems myself, but I wonder why they are calling autoreconf when using the nut-2.6.3 tarball (usually autoreconf is only needed when building from a SVN checkout). Omitting that line from the spec file should help. -- Charles Lepple clepple at gmail
Charles Lepple
2012-Jul-24 12:17 UTC
[Nut-upsuser] Fwd: Re Powercom Vanguard driver for nut
[didn't make it to the list this time, either] Begin forwarded message:> From: Charles Lepple > Subject: Re: [Nut-upsuser] Re Powercom Vanguard driver for nut > Date: July 24, 2012 8:05:59 AM EDT > To: Gordon Longland > Cc: nut-upsuser > > On Jul 22, 2012, at 6:29 PM, Gordon Longland wrote: > >> Compiled 2.6.4 with what I believe to be the right options and can now load the driver usbhid-ups for the powercom 0d9f:0004. >> But just get a rolling screen of > > With at least one -D flag, you will get those debug messages, and the driver won't go into the background. > > [...] >> But when trying to get other components to run keep getting a refusal >> upsc VGD at localhost >> Error: Connection failure: Connection refused >> >> Is there something I am missing? Should it run under the powercom driver or the usbhid-ups? > > 'upsc' is a client for 'upsd' (which talks to the driver), so make sure that 'upsd' has started. > > 'usbhid-ups' is the correct driver if you are using the USB port. (Not sure if this UPS hardware has a serial port as well as USB.) > > I added a note to the 'powercom' driver man page to clarify that it is only for serial (RS-232) connections: > > http://buildbot.networkupstools.org/~buildbot/cayenne/docs/r3689-202/man/powercom.html > > This is how the hardware compatibility list looks so far (filtered to just Powercom USB devices): > > http://buildbot.networkupstools.org/~buildbot/cayenne/docs/r3689-202/website/output/stable-hcl.html?manufacturer=Powercom&connection=USB > > Looks like I need to add to an entry for your particular USB Product ID (0004). > > -- > Charles Lepple > clepple at gmail > > >-- Charles Lepple clepple at gmail