[Nut-upsuser] Blazer_usb Permissions problem: Input/output error

Arnaud Quette aquette.dev at gmail.com
Sat Oct 13 11:34:34 UTC 2012


Hi Paul,

2012/10/11 Paul Barber <p at barbz.com.au>

> Arnaud Quette <aquette.dev <at> gmail.com> writes:
>
> >
> >
> > Hi Paul
> > 2012/8/19 Paul Barber <p <at> barbz.com.au>
> > Hi all,
> > I have an eaton ENV800HA connected for a freebsd 9.0 box with nut 2.6.5
> installed (on the supported list).
> >
> > The UPS is connected using the Blazer_USB driver (not sure how to check
> its
> version).
> >
> > When I kick off upsd it detects the UPS no problems and I get all my
> information:
> >
> > battery.charge: 100battery.voltage: 13.60battery.voltage.high:
> 13.00battery.voltage.low: 10.40battery.voltage.nominal: 12.0device.type:
> upsdriver.name: blazer_usbdriver.parameter.pollinterval:
> 2driver.parameter.port:
> /dev/ugen1.5driver.version: 2.6.5-Unversioned
> directorydriver.version.internal:
> 0.09input.current.nominal: 3.0input.frequency: 50.2input.frequency.nominal:
> 50input.voltage: 245.0input.voltage.fault: 244.5input.voltage.nominal:
> 240output.voltage: 245.0ups.beeper.status: enabledups.delay.shutdown:
> 30ups.delay.start: 180ups.load: 4ups.productid: 5161ups.status:
> OLups.temperature: 25.0ups.type: offline / line interactiveups.vendorid:
> 0665
> >
> > However after a random amount of time I get the following error:
> > blazer_usb[4466]: Permissions problem: Input/output error
> >
> > Followed by:
> > upsd[4468]: Can't connect to UPS [EatonUPS] (blazer_usb-EatonUPS): No
> such
> file or directoryupsmon[4566]: Poll UPS [EatonUPS <at> localhost] failed -
> Driver not connectedupsmon[4566]: Communications with UPS EatonUPS <at>
> localhost lostupssched-cmd: Communications with the UPS EatonUPS <at>
> localhost
> are lostupsmon[4566]: Poll UPS [EatonUPS <at> localhost] failed - Driver
> not
> connected
> >
> > Ive tried the basics and chmod 777'd the ugen1.5 port (not a problem it
> if
> works for a while) but worth a shot.
> >
> > After doing some reading on here I
> ran /usr/local/libexec/nut/./blazer_usb -u
> root -DDD -a EatonUPS which on its first run worked for 72 seconds, but 330
> seconds on the second run before ending with:
> >
> >
> > 327.930857   send: Q1 328.193486   read: (247.0 247.0 247.0 004 50.0
> 13.6 25.0
> 00001001 329.940841   send: Q1 330.177471   read: (247.0 247.0 247.0 004
> 50.0
> 13.6 25.0 00001001 331.950574   send: Q1 332.193457   read: (247.0 247.0
> 247.0
> 004 50.0 13.6 25.0 00001001 338.860148   send: Unknown error 338.860204
>  Permissions problem: Input/output error
> >
> > Ive also found the following message in the log when I start upsd:
> > root: Unknown USB device: vendor 0x0665 product 0x5161 bus uhub2
> >
> >
> >
> > is this msg different from the one when you initially plug your UPS USB
> cord?
> >
> >
> >  Any ideas where to go from here?
> >
> >
> > good question!blazer_usb will reconnect upon certain error, like your IO
> one.but the thing is that device permissions must allow that.
> > Ie, on Linux, udev is in charge of setting the permissions for NUT on all
> known devices.thus, upon any kind of disconnection/reconnection, NUT will
> be
> able to establish again the communication.you should dig around
> this...cheers,Arnaud-- Linux / Unix / Opensource Engineering Expert -
> Eaton -
> http://opensource.eaton.comNetwork UPS Tools (NUT) Project Leader -
> http://www.networkupstools.org
> > Debian Developer - http://www.debian.orgFree Software Developer -
> http://arnaud.quette.fr
> >
> >
> > _______________________________________________
> > Nut-upsuser mailing list
> > Nut-upsuser <at> lists.alioth.debian.org
> > http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/nut-upsuser
>
> Had a bit more of a play with this and still couldn't make it work.
>
> Ive tried this now also on openindiana with the latest nut - same issue
> works
> for 5 minutes then drops out with stale data errors (at the same time as
> the
> permission errors in BSD).
>
> Ive played with the polling times out to 30 seconds and this hasn't made a
> difference.
>
> Also when it does go to stale data it wont also detect again (I had the
> same
> problems on free bsd)
>
> For example:
> paul at indy:/opt/nut/bin# ./blazer_usb -u root -DDD -a trust
> Network UPS Tools - Megatec/Q1 protocol USB driver 0.09 (2.6.5)
>    0.000000     debug level is '3'
>    0.093158     Checking device (0665/5161) (/dev/usb/665.5161/0)
>    0.100375     - VendorID: 0665
>    0.100413     - ProductID: 5161
>    0.100433     - Manufacturer: INNO TECH
>    0.100455     - Product: USB to Serial
>    0.100475     - Serial Number: 20100826
>    0.100493     - Bus: /dev/usb
>    0.100513     Trying to match device
>    0.100533     Device matches
>    0.100601     Trying megatec protocol...
>    0.101337     send: Q1
>    0.351941     read: (248.5 248.5 248.5 003 50.0 13.6 25.0 00001001
>    0.352107     Status read in 1 tries
>    0.352139     Supported UPS detected with megatec protocol
>    0.352710     send: F
>    0.511933     read: #240.0 003 12.00 50.0
>    0.511997     Ratings read in 1 tries
>    0.512731     send: I
>    0.571931     read: I
>    0.571972     blazer_vendor: short reply
>    0.571994     Vendor information read 1 failed
>    0.572876     send: I
>    0.641968     read: I
>    0.642009     blazer_vendor: short reply
>    0.642031     Vendor information read 2 failed
>    0.642748     send: I
>    0.701980     read: I
>    0.702021     blazer_vendor: short reply
>    0.702042     Vendor information read 3 failed
>    0.702066     Vendor information unavailable
>    0.702087     No values provided for battery high/low voltages in
> ups.conf
>    0.702120     Using 'guestimation' (low: 10.400000, high: 13.000000)!
>    0.702142     Battery runtime will not be calculated (runtimecal not set)
>    0.702718     send: Q1
>    0.961973     read: (248.5 248.5 249.0 003 50.0 13.6 25.0 00001001
>    0.962584     dstate_init: sock /var/ups/blazer_usb-trust open on fd 7
>    0.963367     send: Q1
>    1.211876     read: (249.0 248.5 248.5 003 50.0 13.6 25.0 00001001
>    2.962548     send: Q1
>    3.231991     read: (248.5 248.5 248.5 003 50.0 13.6 25.0 00001001
>    4.952462     send: Q1
>    5.211943     read: (248.5 249.0 248.5 003 50.0 13.6 25.0 00001001
>    6.942513     send: Q1
>    7.201996     read: (249.0 248.5 248.5 003 50.0 13.6 25.0 00001001
> ^C
>
> Then after it drops out:
>

I'd like to see the continuation of the trace above, including the drop out
and reconnection attempts.

I would also need the 2 following info:
- what is the actual status of the device file (/dev/usb/665.5161/0), i.e
permissions
- any usb related messages from your system log.
the below "(0000/0000)" for "Checking device" implies that there is a
kernel issue.
it should be as above (0665/5161).


> paul at indy:/opt/nut/bin# ./blazer_usb -u root -DDD -a trust
> Network UPS Tools - Megatec/Q1 protocol USB driver 0.09 (2.6.5)
>    0.000000     debug level is '3'
>    3.972638     Checking device (0000/0000) (/dev/usb/665.5161/0)
>    8.949923     Failed to open device, skipping. (Invalid argument)
>    8.983461     No supported devices found. Please check your device
> availability with 'lsusb'
> and make sure you have an up-to-date version of NUT. If this does not help,
> try running the driver with at least 'subdriver', 'vendorid' and
> 'productid'
> options specified. Please refer to the man page for details about these
> options
> (man 8 blazer).
>
> Any suggestions where to next?


I still don't have a clear visibility on USB with *BSD.
which exact BSD are you using?
what is the result of an "ldd ./blazer_usb"?
there are known issues with the libusb re-implementation on FreeBSD...

Im considering tossing the UPS and buying an APC.
>

understandable.

cheers,
Arnaud
-- 
Network UPS Tools (NUT) Project Leader - http://www.networkupstools.org
Debian Developer - http://www.debian.org
Free Software Developer - http://arnaud.quette.fr
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.alioth.debian.org/pipermail/nut-upsuser/attachments/20121013/da08a3d3/attachment-0001.html>


More information about the Nut-upsuser mailing list