CUPS error messages

I have been unsuccessful in installing a HP6l printer in a suse desktop
stand alone system not attached to the internet or anything else. With the
browser interface and Yast interface as root I get repeated "client error
msg "request_too_long" . With Yast I get that plus "failure_to
communicate_with CUPS" or "CUPS_died during _ initialization" at different
points in the process. The system recognizes the printer and the parallel
port, LP0, checks out OK printing "Hello world" but not the printer test
page or creation of a default printer. The system is useless without the
printer!The configurations have been "cleared" CUPS reloaded etc. to no
avail.

This a system loaned to Mayor Billings by Novell to facilitate production of
campaign letters and to get a little free publicity for suse Linux! The
printer was physically attached but not installed. We're doing the job with
Windows.

> I have been unsuccessful in installing a HP6l printer in a suse desktop
> stand alone system not attached to the internet or anything else. With the
> browser interface and Yast interface as root I get repeated "client error
> msg "request_too_long" . With Yast I get that plus "failure_to
> communicate_with CUPS" or "CUPS_died during _ initialization" at different
> points in the process. The system recognizes the printer and the parallel
> port, LP0, checks out OK printing "Hello world" but not the printer test
> page or creation of a default printer. The system is useless without the
> printer!The configurations have been "cleared" CUPS reloaded etc. to no
> avail.
>
> This a system loaned to Mayor Billings by Novell to facilitate production of
> campaign letters and to get a little free publicity for suse Linux! The
> printer was physically attached but not installed. We're doing the job with
> Windows.
>
>

Looks like youre having the same issue I am, but with a different printer,
so my question is, Is this a Suse problem?
Did Suse update something that conflicts with Cups? or vise versa?

On Thu, 06 Oct 2005 17:17:50 GMT
barrynewton1@comcast.net wrote:

> Looks like youre having the same issue I am, but with a different
> printer, so my question is, Is this a Suse problem?
> Did Suse update something that conflicts with Cups? or vise versa?

Just for another data point, we had trouble making the HP6L (I *think*
it was a 6L, it was a 6something) work via CUPS on Mandrake 9 and 10.
Had to use LPD . . .


--
Kevin Nathan (Arizona, USA)
Linux Potpourri and a.o.l.s. FAQ -- http://www.project54.com/linux/

Open standards. Open source. Open minds.
The command line is the front line.
Linux 2.6.8-24.18-default
9:38pm up 28 days 22:25, 13 users, load average: 0.31, 0.31, 0.30

"Kevin Nathan" <knathan@project54.com> wrote in message
news:20051006214354.601f1dab@efreet.linux...
> On Thu, 06 Oct 2005 17:17:50 GMT
> barrynewton1@comcast.net wrote:
>
> > Looks like youre having the same issue I am, but with a different
> > printer, so my question is, Is this a Suse problem?
> > Did Suse update something that conflicts with Cups? or vise versa?
>
> Just for another data point, we had trouble making the HP6L (I *think*
> it was a 6L, it was a 6something) work via CUPS on Mandrake 9 and 10.
> Had to use LPD . . .
>
>
> --
> Kevin Nathan (Arizona, USA)
> Linux Potpourri and a.o.l.s. FAQ -- http://www.project54.com/linux/
>
> Open standards. Open source. Open minds.
> The command line is the front line.
> Linux 2.6.8-24.18-default
> 9:38pm up 28 days 22:25, 13 users, load average: 0.31, 0.31, 0.30

It's a Suse or CUPS problem. Printer is a HP Laserjet 6l and I actually had
it working OK when frist installed using Yast but somehow it got screwed up
and I keep getting the error messages mentioned which are anything but
intuitive. I suspect that in some obscure configuration file there is a
limit on the string length causing at least part of the problem.

On Fri, 07 Oct 2005 17:26:28 GMT
"Jim Monnahan" <jrmonna@comcast.net> wrote:

> It's a Suse or CUPS problem. Printer is a HP Laserjet 6l and I
> actually had it working OK when frist installed using Yast but
> somehow it got screwed up and I keep getting the error messages
> mentioned which are anything but intuitive. I suspect that in some
> obscure configuration file there is a limit on the string length
> causing at least part of the problem.
>
>

You might want to follow this procedure to do a complete un-install and
re-install of CUPS:

http://portal.suse.com/sdb/en/2004/0...reinstall.html

If that doesn't help, let us know -- and give us the *exact* error
msg(s), along with any other data points you think are pertinent . . .


--
Kevin Nathan (Arizona, USA)
Linux Potpourri and a.o.l.s. FAQ -- http://www.project54.com/linux/

Open standards. Open source. Open minds.
The command line is the front line.
Linux 2.6.8-24.18-default
12:51am up 30 days 1:38, 13 users, load average: 0.32, 0.35, 0.35

I've considered reinstalling the whole system but it's not my machine and
I'm a little reluctant to mess with it and/or directly modify the config
files.

With the web interface the printer appears and ready to accept but when the
printer test is run I get " the request value is too large for this server
to process"
When I try to reconfigure the printer on the web interface get, "server
error - internal error" which would suggest the problem is with the server
definition or configuration rather than CUPS.

On Yast interface when trying to test printer get "lpr: unable to print
file: client_error_request_value_too_long".
On Yast do get "hello world" printed when checking the parallel port and it
does recognize the printer make and model OK but further printer testing
fails with the above error msg.

On the command line interface I have been able to "echo "phrase" > lp0" and
"ls -l > lp0" and get first line output on the printer apparently because
the printer feed stops at the first <linefeed>.

I suspect the problem is with some obscure entry in the client.conf or
cupsd.conf files. I've tried clearing the configuration on Yast in the
"Advanced" option and also reloading CUPS from Yast and starting from
scratch installing the printer but get the same error.messages when testing.

On Sat, 08 Oct 2005 16:34:06 GMT
"Jim Monnahan" <jrmonna@comcast.net> wrote:

> I've considered reinstalling the whole system but it's not my machine
> and I'm a little reluctant to mess with it and/or directly modify the
> config files.
>

A wise course of action! I didn't realize this was *not* your box.


> With the web interface the printer appears and ready to accept but
> when the printer test is run I get " the request value is too large
> for this server to process"
> When I try to reconfigure the printer on the web interface get,
> "server error - internal error" which would suggest the problem is
> with the server definition or configuration rather than CUPS.
>
> On Yast interface when trying to test printer get "lpr: unable to
> print file: client_error_request_value_too_long".

This message more than likely means: there is not enough space in
/var/spool/cups to create the print image, on the machine from which
you're trying to print (and *not* the machine with the printer attached)

You might need to delete old log files and such from the /var system.


> On Yast do get "hello world" printed when checking the parallel port
> and it does recognize the printer make and model OK but further
> printer testing fails with the above error msg.
>

This is still consistent with a nearly full partition. The small file
prints ok, but the bigger ones don't. See what the output is of:

df -h

and if the partition containing /var is very full, you can do:

su -c "du -ah --max-depth=1 /var | egrep '^[0-9]*[GM] *' | less"

Replace the '/var' with whatever directory in which you want to start.
If you want to see farther down the tree, leave out the '--max-depth=1'
parameter (or increase the number).


> On the command line interface I have been able to "echo "phrase" >
> lp0" and "ls -l > lp0" and get first line output on the printer
> apparently because the printer feed stops at the first <linefeed>.
>
> I suspect the problem is with some obscure entry in the client.conf or
> cupsd.conf files. I've tried clearing the configuration on Yast in the
> "Advanced" option and also reloading CUPS from Yast and starting from
> scratch installing the printer but get the same error.messages when
> testing.
>

If it's not the disk being too full, we'll explore some of these
settings . . . :-)


--
Kevin Nathan (Arizona, USA)
Linux Potpourri and a.o.l.s. FAQ -- http://www.project54.com/linux/

Open standards. Open source. Open minds.
The command line is the front line.
Linux 2.6.8-24.18-default
10:58am up 30 days 11:45, 13 users, load average: 0.45, 0.40, 0.39

"Kevin Nathan" <knathan@project54.com> wrote in message
news:20051008111402.1f50acd9@efreet.linux...
> On Sat, 08 Oct 2005 16:34:06 GMT
> "Jim Monnahan" <jrmonna@comcast.net> wrote:
>
> > I've considered reinstalling the whole system but it's not my machine
> > and I'm a little reluctant to mess with it and/or directly modify the
> > config files.
> >
>
> A wise course of action! I didn't realize this was *not* your box.
>
>
> > With the web interface the printer appears and ready to accept but
> > when the printer test is run I get " the request value is too large
> > for this server to process"
> > When I try to reconfigure the printer on the web interface get,
> > "server error - internal error" which would suggest the problem is
> > with the server definition or configuration rather than CUPS.
> >
> > On Yast interface when trying to test printer get "lpr: unable to
> > print file: client_error_request_value_too_long".
>
> This message more than likely means: there is not enough space in
> /var/spool/cups to create the print image, on the machine from which
> you're trying to print (and *not* the machine with the printer attached)
>
> You might need to delete old log files and such from the /var system.
>
>
> > On Yast do get "hello world" printed when checking the parallel port
> > and it does recognize the printer make and model OK but further
> > printer testing fails with the above error msg.
> >
>
> This is still consistent with a nearly full partition. The small file
> prints ok, but the bigger ones don't. See what the output is of:
>
> df -h
>
> and if the partition containing /var is very full, you can do:
>
> su -c "du -ah --max-depth=1 /var | egrep '^[0-9]*[GM] *' | less"
>
> Replace the '/var' with whatever directory in which you want to start.
> If you want to see farther down the tree, leave out the '--max-depth=1'
> parameter (or increase the number).
>
>
> > On the command line interface I have been able to "echo "phrase" >
> > lp0" and "ls -l > lp0" and get first line output on the printer
> > apparently because the printer feed stops at the first <linefeed>.
> >
> > I suspect the problem is with some obscure entry in the client.conf or
> > cupsd.conf files. I've tried clearing the configuration on Yast in the
> > "Advanced" option and also reloading CUPS from Yast and starting from
> > scratch installing the printer but get the same error.messages when
> > testing.
> >
>
> If it's not the disk being too full, we'll explore some of these
> settings . . . :-)

The df command shows total size of 9.3G with 2.0 used and 7.4 available so I
don't think it's a simple disk full situation. I cleaned out some old log
files that looked big and dropped some users but it didn't make any
difference in the error messages when attempting to print.I can echo
"phrase" to /dev/lp0 and get printer output when I try to redirect a command
like ls -l or cat a file to /dev/lp0 I get just the first line of output.
Apparently the system interprets the first LF as EOF and that's all that
prints even though the same commands not redirected show up fine on the
terminal.Error messages are as initially described. for either Yast or the
browser interface.

The machine from which I'm trying to print is the one to which the printer
is attached. It's just the desktop computer + printer with nothing else
hooked up at the moment.

On Mon, 10 Oct 2005 01:38:58 GMT
"Jim Monnahan" <jrmonna@comcast.net> wrote:

> The df command shows total size of 9.3G with 2.0 used and 7.4
> available so I don't think it's a simple disk full situation.

No, doesn't look like it! :-)


> I cleaned out some old log files that looked big and dropped some
> users but it didn't make any difference in the error messages when
> attempting to print.I can echo "phrase" to /dev/lp0 and get printer
> output when I try to redirect a command like ls -l or cat a file
> to /dev/lp0 I get just the first line of output. Apparently the
> system interprets the first LF as EOF and that's all that prints even
> though the same commands not redirected show up fine on the
> terminal.

Instead of simple redirection, try a command like this:

lpr -P printername nameoffiletoprint


> The machine from which I'm trying to print is the one to which the
> printer is attached. It's just the desktop computer + printer with
> nothing else hooked up at the moment.
>

Have you tried a 'Repair' yet? If not, boot with the install media and
select 'Install' as if doing a new install. After the first few
questions, you will get to a dialog box where one of the options is
something like:

Repair installed system

Select the 'auto' (or something like that) and let it repair any errors
it finds.

If that doesn't help, try removing all CUPS related stuff and install
the LPRNG (or is it LPD?) system, see if you can get it working that
way. If so, you can either uninstall it and try CUPS again, or use it
as is.

If you retry the CUPS system, do a YOU after installing CUPS and before
trying to setup any printers -- just in case.


--
Kevin Nathan (Arizona, USA)
Linux Potpourri and a.o.l.s. FAQ -- http://www.project54.com/linux/

Open standards. Open source. Open minds.
The command line is the front line.
Linux 2.6.8-24.18-default
9:49pm up 32 days 22:36, 13 users, load average: 0.22, 0.10, 0.09

I explored the error logs and found a constant "unable to creat or open
"var/spool.cups" msg so looked at the file system and there wasn't any
folder by this name so I created one. At least now CUPS seems to initialize
OK and the printer installs without crashing either with the web interface
or Yast. On test I get some reaction from the printer, lights flash and or
prints blank pages but no print output. On the web interface the printer
appears OK as default on /dev/lp0, ready to accept etc but still a test page
won't print although sometimes a bunch of blank pages spit out.

Again redirecting like lpstat -t > /dev/lp0 gets printer output but it
appears the printer is recognizing NL's but not LF's in that the multiple
short lines resulting from the command show the first line at the top, then
a NL starting on the next line below the last character of the first line
and so on until the output runs off the page and then stops. It never resets
to the left margin between lines. No doubt there has to be some way to tell
the printer that a LF is to be associated with each NL.

Trying lpr -P as you suggested produced error log msg's as follows:
"started filter usr/lib/cups/filter/testtops (PID6073)"
"started filter /usr/lib/cups/filter /pstops (PID6073)"
"started backend usr/lib/cups/backend/parallel(PID6073)"
"PID 6173 stopped with status 3"
"hint try setting LogLevel to 'debug' to find out more"
I intend to do this when I get another shot at it and figure out how to do
it from the "man set" page.

The Yast interface is not exactly intuitive to me anyway. For example when
installing the printer on a stand alone desktop system not hooked to any
network etc., should one choose the "parallel printer" or the "CUPS Full
Server" configuration? Should one choose "local filtering" or not in this
situation? When choosing "parallel printer" I wind up with a "raw printer".
In this situation I'd guess that the PostScript filter in CUPS is not
functional so that the raw printer wouldn't work with the Suse Open Office
applications. In the CUPS.org documentation it appears that one should
install a full network installation even if it's one stand alone one
computer/one printer setup. In the configuration window the two are mutually
exclusive. I did set up the LRng print system which tested OK but then the
Suse OpenOffice applications wouldn't run - not just wouldn't print but
would hang while starting up!

Years ago I managed to set up a UNIX system with a serial line printer and
no graphic interface to make things "easy" but this Suse/CUPS system has me
buffaloed. Fun and challenging though!