[Nut-upsuser] My previous post (lengthy, lots of tarace output)

[Gene Heskett]

On Thursday 31 December 2009, Arjen de Korte wrote:
>Citeren Gene Heskett <gene.heskett at gmail.com>:
>> You requested logs, which I sent, and which were not commented on in any
>> reply I received.  Those logs show the usbhid-ups/belkin driver
>> combination losing comm with the ups and having to redo its search for
>> the ups several times a minute, but you apparently sent them to
>> /dev/null.  Were I the coder for that piece of code, that would have
>> raised all sorts of red flags.
>
>OK, here you go. The following occurs several times in this log:
>
>     Got disconnected by another driver: Device or resource busy
>
>This means you've been running multiple instances of usbhid-ups in
>parallel, which I explicitly told you not to do. You can expect this
>kind of a mess then, which makes the remainder of the log useless.
>
I saw that, killed all of them and restarted just one instance, which was 
what was running when I made that log.  My logs are being spammed right now 
by only one instance of usbhid-ups running.
-------------------------------------
Dec 31 08:45:47 coyote upsd[888]: Data for UPS [myups] is stale - check 
driver
Dec 31 08:45:58 coyote upsd[888]: UPS [myups] data is no longer stale
Dec 31 08:46:21 coyote upsd[888]: Data for UPS [myups] is stale - check 
driver
Dec 31 08:46:50 coyote upsd[888]: UPS [myups] data is no longer stale
Dec 31 08:46:55 coyote upsd[888]: Data for UPS [myups] is stale - check 
driver
Dec 31 08:47:28 coyote upsd[888]: UPS [myups] data is no longer stale
Dec 31 08:47:47 coyote upsd[888]: Data for UPS [myups] is stale - check 
driver
Dec 31 08:48:02 coyote upsd[888]: UPS [myups] data is no longer stale
Dec 31 08:48:09 coyote upsd[888]: Data for UPS [myups] is stale - check 
driver
Dec 31 08:48:30 coyote upsd[888]: UPS [myups] data is no longer stale
Dec 31 08:48:47 coyote upsd[888]: Data for UPS [myups] is stale - check 
driver
Dec 31 08:49:40 coyote upsd[888]: UPS [myups] data is no longer stale
Dec 31 08:49:53 coyote upsd[888]: Data for UPS [myups] is stale - check 
driver
-----------------------------
[root at coyote grub]# ps -ae|grep ups
  766 ?        00:00:12 usbhid-ups
  888 ?        00:00:00 upsd
 3136 ?        00:00:00 cupsd
----------------------------
Humm, I was just reminded by something I wrote on another ML, I have an 8Gb 
usbkey plugged in, and quite a few other usb devices that connect as storage 
devices.  Storage devices, and the latencies caused by scanning them, can and 
have totally wrecked parts being cut by my cnc milling machine.  I wonder if 
that same mechanism is at work here.  So I'll unplug everything 'storage' 
except the printer, an All-In-one which has a (&*% card reader in it.  It may 
have changed, but didn't fix.

Here is a little more of the log, while I had upsmon running for a short 
time:
----------------------------------------------
Dec 31 10:19:37 coyote upsd[888]: User ups at 127.0.0.1 logged into UPS [myups]
Dec 31 10:19:37 coyote upsmon[12804]: Master privileges unavailable on UPS 
[myups at localhost]
Dec 31 10:19:37 coyote upsmon[12804]: Response: [ERR ACCESS-DENIED]
-------------------------------------------------

Huh?  It is _not_ reporting that one-screen. A clue!

-------------------------------------------------
Dec 31 10:19:42 coyote upsmon[12804]: Poll UPS [myups at localhost] failed - 
Data stale
Dec 31 10:19:42 coyote upsmon[12804]: Communications with UPS myups at localhost 
lost
Dec 31 10:19:42 coyote wall[12807]: wall: user ups broadcasted 1 lines (46 
chars)
Dec 31 10:19:47 coyote upsmon[12804]: Poll UPS [myups at localhost] failed - 
Data stale
Dec 31 10:19:47 coyote upsmon[12804]: UPS myups at localhost is unavailable
Dec 31 10:19:47 coyote wall[12816]: wall: user ups broadcasted 1 lines (36 
chars)
Dec 31 10:19:52 coyote upsmon[12804]: Poll UPS [myups at localhost] failed - 
Data stale

[...]

Dec 31 10:21:22 coyote upsmon[12804]: Poll UPS [myups at localhost] failed - 
Data stale
Dec 31 10:21:23 coyote upsmon[12804]: Signal 15: exiting
Dec 31 10:21:23 coyote upsd[888]: User ups at 127.0.0.1 logged out from UPS 
[myups]
----------------------------------------
Should I be using "slave" in the MONITOR line?  Or should I just gzip the 
.conf dir and send it?

>[...]
>
>> The instructions in the example files are even occasionally miss-leading,
>> and I just found from an error message, that the allowfrom has been
>> deprecated but not removed from one of the .conf files.
>
>True, and this is already fixed in the development version and will
>only log an informational message, nothing more, nothing less. It's
>this kinds of comments that bothers me the most in your replies and
>which makes it harder and harder for me to start reading yet another
>reply from you.

Sorry Arjen, but to me, not knowing where your bugzilla might be, I look at 
that as a bug report, and its certainly not intended to be a personal attack.

>[...]
>
>>> I'm sorry, but the best advice I can give you here is to
>>> find an RPM where this is already done for you.
>>
>> Which I have explained is not available for F10.
>
>The rebuild the SRPMS for a newer version. For someone with so much
>experience, I would have expected that you'd be able to know that this
>is usually the easiest way to get a newer version of some package
>supported in an older distro. At the very least, it might give you a
>head start for finding out which configure flags are needed in Fedora
>10 since they will be listed in the .spec file.

And that is one thing I have not mastered, the use of rpmbuild.  If the .spec 
file needs massaged, I am clueless.  I think I might have made it work once.

>Best regards, Arjen

Thank you Arjen.  If you ever make it to this side of the pond and thence to 
West Virginia, I'm sure I could find a hand cooler in the fridge.

-- 
Cheers, Gene
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)

The only recomendation is "dont".

	- Alan Cox giving some recommendations for binary-only drivers