Documents silently fail to print

[Bodinux]

Binary package hint: cups

ubuntu 9.04 updated

I am trying to print a simple .pnm file with eog on the magicolor 2530 dl printer but nothing comes out.

When I try to print a test page, it is successful.

No problem at all with my usb Brother printer.

ProblemType: Bug
Architecture: i386
DistroRelease: Ubuntu 9.04
Lpstat:
 périphérique pour HL-5040-series : usb://Brother/HL-5040%20series
 périphérique pour magicolor-2530-DL : socket://192.168.1.117:9100
MachineType: ASUSTeK Computer INC. A7N8X-E
Package: cups 1.3.9-17ubuntu3
Papersize: a4
PpdFiles:
 magicolor-2530-DL: KONICA MINOLTA magicolor 2530 DL Foomatic/foo2lava (recommended)
 HL-5040-series: Brother HL-5040 Foomatic/hl1250 (recommended)
ProcCmdLine: root=UUID=c6ea8388-f725-4a21-af69-3ec86ae7d4f0 ro quiet splash
ProcEnviron:
 LANG=fr_FR.UTF-8
 SHELL=/bin/bash
ProcVersionSignature: Ubuntu 2.6.28-11.42-generic
SourcePackage: cups

[Bodinux (bodinux)]

Here is the attachement of the /var/log/cups/error-log file

[Till Kamppeter (till-kamppeter)]

The only jobs which got captured in the error_log in debug mode a successfully printed test pages. The failed eog jobs are captured but not in debug mode. Please follow the instructions in the section "CUPS error_log" on https://wiki.ubuntu.com/DebuggingPrintingProblems and print jobs from eog to get them captured.

[Bodinux (bodinux)]

okay, here is the new error_log.

Job10 starts at : 15:15:06

it seems that gs runs until 15:29:19 (which is not really acceptable in itself...)

and then there is a broken pipe error.

Jobs is still going on anyhow

Then I print a test page, job 11

When I cancel the previous job a few minutes later, the test page does print.

[Till Kamppeter (till-kamppeter)]

This is most probably a driver bug, as the job prints correctly on your Brother printer. Therefore the application and the CUPS filters seem to do their part correctly.

So the problem is probably that the driver (foo2lava, package: foo2zjs) sends something to the printer which makes it crash or block so that it does not accept further data. The driver itself does not crash, it converts the complete job and foomatic-rip closes successfully. The backend closes with an error before the the end of the job is reached as the printer stops accepting further data.

Moving the bug to the foo2zjs package ...

[Till Kamppeter (till-kamppeter)]

Can you print from other applications, like Evince, OpenOffice.org, or Firefox? Does this work?

[Till Kamppeter (till-kamppeter)]

Rick, can you have a look into this?

[Rick Richardson (rick-richardson)]

$ eog /usr/share/sane/xsane/xsane-startimage.pnm
Print -> 2530dl -> Advanced -> Color Mode -> "Color - use ICM color profile"
ICM Color Profile -> "km2530 (default)"
-> Print

Works good on Fedora 10.

Works good on Ubuntu 9.04 as released.

Works good on Ubuntu 9.04 current.

[Bodinux (bodinux)]

Okay, it works here also for me.

So I did the test again with my picture and your settings and the picture finally printed more than 35 minutes later !

Then I have done the test again with these settings :
Color Mode -> "Color - use ICM color profile"
ICM Color Profile -> "File user1.icm"

This one did not print.

[Bodinux (bodinux)]

Attached is the image file that doesn't print (simple scan with xsane)

[Till Kamppeter (till-kamppeter)]

According to the error_log it seems that the CUPS backend ("socket" here) only connects to the printer once in a while. Can you try whether the printer works better with the LPD or IPP CUPS backends?

[Bodinux (bodinux)]

Is it possible that the ghostscript takes so much time to complete that
the job 18 (see error_log) is removed before competion ?

I need more info to do what you ask me as I have no clue on how to do
it. (LPD or IPP CUPS backend).

I wonder if Rick was able to print the file.

[Rick Richardson (rick-richardson)]

$ eog ./out.pnm

Printed in ~30 seconds from printer up to printer down on Fedora 10. Sorry, works great here.

[Till Kamppeter (till-kamppeter)]

Bodinux, it is not Ghostscript or the driver which takes so much time. It seems that the printer fails to connect in most cases. On a failed attempt CUPS retries some minutes later, up to 5 times. So if you get your job printed after 35 minutes, it is because some attempts to connect failed and once an attempt succeeded, 35 minutes after submitting the job. If you do not get a printout at all, 5 attempts failed and CUPS gave up.

To set up a printer with another than the standard backend (LPD, IPP, SMB) start system-config-printer via

System -> Administration -> Printing

Click the "New" button. The New Printer wizard pops up. There click the triangle at "Network Printer" to expand the list for the network printers.

For the LPD backend

Choose "LPD/LPR Host or Printer". Enter the IP address of your printer into the "Host" field on the right. Leave the "Queue" field blank and do not click on "Probe". Click on "Forward". Now choose the manufacturer, click "Forward" and choose the model. Ass driver you will probably only get one choice, the foo2lava driver. Click "Forward". Edit the fields if you want, and click "Apply". Click "Yes" for printing a test page. Does it print? If not, right-click the new printer icon and check the options, especially the page size.

For the IPP backend

Choose "Internet Printing Protocol". Enter the IP address of your printer into the "Host" field on the right. Click the "Find queue ..." button. Choose a queue and click "Forward". Now choose the manufacturer and model and do the rest as described above for the LPD backend.

For the SMB backend

Choose "Windows printer via SAMBA". Click "Browse" on the right. If the printer appears as one host in the list, choose it. If not, it does not support SMB or it is not activated.

Check also whether LPD, IPP, and/or SMB access on the printer is active by the printer's administration interface. Access the printer's administration interface by entering the printer's IP address into the URL field of a web browser.

[Bodinux (bodinux)]

Okay for your explanation, nevertheless my cpu is up to 100% with gs and
sed during all the tries to connect. I took this as a sign that the
processing was not yet terminated.

Regarding my tries :

LPD :
*supported by the printer
*I never achieved to print anything (even the test page)
*URI after following your indications : lpd://192.168.1.117

SMB :
*supported by the printer
*I had to install 'samba'
*After waiting several minutes, the 'workgroup' finally appeared at
'browse...' and I could find the printer
*test page prints on two pages (same result on two tries):
 **one page with : %-12345X@PJL JOB NAME="stdin"
 **one page almost entirely black with some vertical stripes (monochrome
printing)
*see [job 41] vas a color test page

IPP :
*supported by the printer
*I had to fiddle with the URI to get some results : URI:
ipp://192.168.1.117:80/ipp
*works well
*[job 42] the out.pnm file with Icm color corection : default.icm
*[job 43] the out.pnm file with Icm color correction : file 1.icm (the
one which causes problem)

In short :
No success in lpd : printer not seen (bad URI probably)
No success in smb : garbage sent to printer (bug ?)
Success with ipp which prints the file in less than 6 minutes

Conclusion :
Alhtough it takes an awful lot of time, the printer is now printing this
image.
At least two problem remain :
*the socket problem
*the samba problem

If you feel there is a need to continue working on these items, I'll be
glab to help if I can.

Thanks for your help.

[Bodinux (bodinux)]

error_log containing job debug

[Till Kamppeter (till-kamppeter)]

Can you please attach an uncompressed error_log? Thanks.

[Till Kamppeter (till-kamppeter)]

As the document prints correctly with the printer connected via IPP the driver must be correct. So moving from foo2zjs to CUPS ...

[Rick Richardson (rick-richardson)]

I think he has a network problem... 5MB should not take 6 minutes.

[equivalent of lpr]

$ time foo2lava-wrapper -c output.ps > xxx

real 0m36.191s
user 0m39.563s
sys 0m0.456s
$ ll xxx
-rw-rw-r-- 1 rick rick 5057183 2009-05-13 09:46 xxx
$ nc 192.168.1.13 9100 <xxx

[Bodinux (bodinux)]

Rick, can you please show me the commands to cut the printing process
into time measurable pieces ?

My network is rather simple and I achieve high transfert rates between
my elements, moreover printing from windows takes no time.

[Bodinux (bodinux)]

error_log unzipped

[Rick Richardson (rick-richardson)]

$ eog ./out.pnm
(print to a file: output.ps)

$ foo2lava-wrapper -c -p [paper] output.ps > xxx
... [paper] defaults to letter...

$ nc 192.168.1.13 9100 < xxx

See:

$ man foo2lava-wrapper
$ man nc

$ grep foo2lava /var/log/messages

I DON"T do CUPS error logs!!! Period.

This will be my last post. Please post on:

http://foo2zjs.rkkda.com/forum/index.php

[Bodinux (bodinux)]

Well, I did as Rick mentionned

eog prints in no time into output.ps

foo2lava-wrapper takes 43 seconds (which is okay compared to the 39 seconds of Rick)

and the printer takes no time to receive the data and to print it. (->no problem with my network)

I try to print from GIMP, and it takes a looong time again (26 minutes) but finally prints (see error_log _44 for details)

[Bodinux (bodinux)]

Till

What is necessary to change the status of this bug ?

[Till Kamppeter (till-kamppeter)]

According to your error_log the job was executed repeatedly, until the CUPS backend once succeeded to connect tou your printer. Then the job went out to the printer in no time (or at least a reasonable time).

If you click on the state of the bug some interface for changing it (and other things like the package to which the bug is assigned) appears. If this interface does not appear for you or if the items you want to change are blocked, you have no permissions to change these item. Please tell in a comment how you want to have the state items changed.

[Bodinux (bodinux)]

Do we have enough information to set the bug status to confirmed ?

If no, what is missing ?

[papukaija (papukaija)]

Thank you for taking the time to report this bug and helping to make Ubuntu better. We are sorry that we do not always have the capacity to look at all reported bugs in a timely manner. There have been many changes in Ubuntu since that time you reported the bug and your problem may have been fixed with some of the updates. If you could test the current Ubuntu development version, this would help us a lot. If you can test it, and it is still an issue, we would appreciate if you could upload updated logs by running apport-collect 374416, and any other logs that are relevant for this particular issue. Please feel free to report any other bugs you may find.

[Charlie Kravetz (cjkgeek)]

We'd like to figure out what's causing this bug for you, but we haven't heard back from you in a while. Could you please provide the requested information? Thanks!

[Bodinux (bodinux)]

I did the test again, with the latest ubuntu 10.10 devel from today (2010.10.01)
Printer URI : socket://192.168.xxx.xxx:9100

$ eog ./out.pnm
Print -> 2530dl -> Advanced -> Color Mode -> "Color - use ICM color profile"
ICM Color Profile -> "km2530 (default)"
-> Print

It printed in a normal time

I also tried using the profile "file user1.icm"

Same good result.

So I guess that it is not a bug anymore !

Thanks to all.

[papukaija (papukaija)]

This bug report is being closed due to your last comment regarding this being fixed with an update. For future reference you can manage the status of your own bugs by clicking on the current status in the yellow line and then choosing a new status in the revealed drop down box. You can learn more about bug statuses at https://wiki.ubuntu.com/Bugs/Status. Thank you again for taking the time to report this bug and helping to make Ubuntu better. Please submit any future bugs you may find.

[Bodinux (bodinux)]

I updated my computers to 10.10 in order to benefit from the resolution of this bug.

Sadly, this bug is still valid. There is no problem to print the Ubuntu test page, but a simple OpenOffice document will simply die between the network card and the printer.

[Bodinux (bodinux)]

More news on this item:

The above problems are fixed : printing from the network is not a problem anymore (as per my test from this morning).

Nevertheless if I have a two page document, with one page in color and the other in B&W, the document doesn't print with the color queue. As it says in the title of this bug : it silently fails to print.

[papukaija (papukaija)]

There should be now enough information in this bug.

[Till Kamppeter (till-kamppeter)]

Bodinux, can you attach the document? Which application did you use? Can you also attach an error_log ("CUPS error_log" section of https://wiki.ubuntu.com/DebuggingPrintingProblems)? Can you also capture what is actually going to the printer ("Capturing print job data" section of https://wiki.ubuntu.com/DebuggingPrintingProblems)?

[Bodinux (bodinux)]

I did as you asked.

make a two page file (one page with color, the second page only black and white) with open office, then obtain a pdf from it.

Open the pdf file with Document Viewer 2.32.0 and follow the 'Capturing job data' procedure while printing it to the colorQueue of the magicolor2530DL through USB.

Trace from terminal:
:~$ cancel -a
:~$ cupsdisable magiCouleur
:~$ lpstat -o
magiCouleur-18 me 14336 mer. 13 oct. 2010 20:58:14 CEST
:~$ sudo ls -1 /var/spool/cups
[sudo] password for me:
c00018
d00018-001
tmp
:~$ sudo cp /var/spool/cups/d00018-001 ~/printout
:~$ sudo chmod 777 /home/laurent/printout
:~$ cupsenable magiCouleur
:~$ file printout
printout: PDF document, version 1.5
:~$

Nothing new in cups/error.log
Some traces in access_log
   localhost - - [13/Oct/2010:20:57:00 +0200] "POST /admin/ HTTP/1.1" 200 164 Purge-Jobs successful-ok
   localhost - - [13/Oct/2010:20:57:10 +0200] "POST / HTTP/1.1" 200 255 Create-Printer-Subscription successful-ok
   localhost - - [13/Oct/2010:20:57:35 +0200] "POST /admin/ HTTP/1.1" 401 159 Pause-Printer successful-ok
   localhost - me [13/Oct/2010:20:57:35 +0200] "POST /admin/ HTTP/1.1" 200 159 Pause-Printer successful-ok
   localhost - - [13/Oct/2010:20:58:14 +0200] "POST /printers/magiCouleur HTTP/1.1" 200 14696 Print-Job successful-ok
   localhost - - [13/Oct/2010:20:58:15 +0200] "POST / HTTP/1.1" 200 344 Create-Printer-Subscription successful-ok
   localhost - - [13/Oct/2010:21:00:08 +0200] "POST /admin/ HTTP/1.1" 401 159 Resume-Printer successful-ok
   localhost - me [13/Oct/2010:21:00:08 +0200] "POST /admin/ HTTP/1.1" 200 159 Resume-Printer successful-ok
Two lines in cups/page.log
   magiCouleur 18 me [13/Oct/2010:21:00:10 +0200] 1 1 - localhost print_fail.pdf iso_a4_210x297mm -
   magiCouleur 18 me [13/Oct/2010:21:00:11 +0200] 2 1 - localhost print_fail.pdf iso_a4_210x297mm -

Attachments:
-pdf file
-printout file

[Bodinux (bodinux)]

The author of the foo2lava driver recommends the use of two queues : one color queue and one black&white queue.

The color page alone prints correctly on the color queue.

The B&W page prints correctly on the B&W queue.

If the two pages are printed on the color queue, they are sent to the printer which warms up and nothing prints.

[Bodinux (bodinux)]

Here is the input of the driver developper :

http://foo2zjs.rkkda.com/forum/read.php?21,2773

"
$ lpr -P mc2530c ~/tmp/print_fail.pdf

$ lpr -P mc2530 ~/tmp/print_fail.pdf

Works fine on Fedora 13 in mono OR color.
"

The same command doesn't work on Ubuntu 10.10. The printer goes into preheating then nothing.

[Rick Richardson (rick-richardson)]

From the Changelog:

2010-10-16 Rick Richardson <email address hidden>
        * foo2lava.c: Iff MODE_COLOR and nbie = 1, then it is a mono page.

Please download, make, make install again.

[Bodinux (bodinux)]

Great improvment : the two pages now print on the two queues!

Nevertheless, on the black and white queue, the first page also prints in color. I was expecting it to print in B&W.

[Launchpad Janitor (janitor)]

[Expired for cups (Ubuntu) because there has been no activity for 60 days.]