Bug 1050607 - dnssd failure with IPv6
Summary: dnssd failure with IPv6
Keywords:
Status: CLOSED EOL
Alias: None
Product: Fedora
Classification: Fedora
Component: cups
Version: 20
Hardware: x86_64
OS: Unspecified
unspecified
unspecified
Target Milestone: ---
Assignee: Tim Waugh
QA Contact: Fedora Extras Quality Assurance
URL: https://retrace.fedoraproject.org/faf...
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2014-01-08 23:31 UTC by collura
Modified: 2015-06-29 14:16 UTC (History)
3 users (show)

Fixed In Version:
Clone Of:
Environment:
Last Closed: 2015-06-29 14:16:57 UTC
Type: ---
Embargoed:


Attachments (Terms of Use)
File: backtrace (25.21 KB, text/plain)
2014-01-08 23:32 UTC, collura
no flags Details
File: cgroup (155 bytes, text/plain)
2014-01-08 23:32 UTC, collura
no flags Details
File: core_backtrace (5.16 KB, text/plain)
2014-01-08 23:32 UTC, collura
no flags Details
File: dso_list (3.86 KB, text/plain)
2014-01-08 23:32 UTC, collura
no flags Details
File: environ (896 bytes, text/plain)
2014-01-08 23:32 UTC, collura
no flags Details
File: exploitable (82 bytes, text/plain)
2014-01-08 23:32 UTC, collura
no flags Details
File: limits (1.29 KB, text/plain)
2014-01-08 23:32 UTC, collura
no flags Details
File: maps (24.03 KB, text/plain)
2014-01-08 23:32 UTC, collura
no flags Details
File: open_fds (1.06 KB, text/plain)
2014-01-08 23:32 UTC, collura
no flags Details
File: proc_pid_status (900 bytes, text/plain)
2014-01-08 23:32 UTC, collura
no flags Details
File: var_log_messages (4.52 KB, text/plain)
2014-01-08 23:32 UTC, collura
no flags Details
clog job (75.41 KB, application/octet-stream)
2014-06-18 08:25 UTC, collura
no flags Details
snippet of var log messages near time of print que clog during rapid addition of jobs (7.05 KB, text/plain)
2014-06-18 10:20 UTC, collura
no flags Details
part of the var log cups error_log for comment#19 but also has old gs snipet (441.44 KB, text/plain)
2014-06-18 13:08 UTC, collura
no flags Details

Description collura 2014-01-08 23:31:56 UTC
Description of problem:
often when i print multiple sequential jobs, only the first one makes it through.  seems that if i forget and reboot that the old jobs come through then.  hoping this crash might be the issue so can get it fixed.  have had this in fc19 as well as fc20.

Version-Release number of selected component:
ghostscript-9.10-5.fc20

Additional info:
reporter:       libreport-2.1.10
backtrace_rating: 4
cmdline:        gs -q -dNOPAUSE -dBATCH -dSAFER -sDEVICE=ps2write -sOUTPUTFILE=%stdout -dLanguageLevel=3 -r1200 -dCompressFonts=false -dNoT3CCITT -dNOINTERPOLATE -c 'save pop' -f /var/spool/cups/tmp/cupsscNQCL
crash_function: gx_device_enum_ptr
executable:     /usr/bin/gs
kernel:         3.13.0-0.rc6.git0.1.fc21.x86_64
runlevel:       N 5
type:           CCpp
uid:            4

Truncated backtrace:
Thread no. 1 (10 frames)
 #0 gx_device_enum_ptr at base/gsdevice.c:97
 #1 pdf14_device_enum_ptrs at base/gdevp14.c:496
 #2 gc_trace at psi/igc.c:857
 #3 gs_gc_reclaim at psi/igc.c:335
 #4 context_reclaim at psi/zcontext.c:280
 #5 gs_vmreclaim at psi/ireclaim.c:155
 #6 ireclaim at psi/ireclaim.c:77
 #7 interp_reclaim at psi/interp.c:441
 #8 interp at psi/interp.c:1713
 #9 gs_call_interp at psi/interp.c:510

Comment 1 collura 2014-01-08 23:32:05 UTC
Created attachment 847387 [details]
File: backtrace

Comment 2 collura 2014-01-08 23:32:11 UTC
Created attachment 847388 [details]
File: cgroup

Comment 3 collura 2014-01-08 23:32:12 UTC
Created attachment 847389 [details]
File: core_backtrace

Comment 4 collura 2014-01-08 23:32:15 UTC
Created attachment 847390 [details]
File: dso_list

Comment 5 collura 2014-01-08 23:32:20 UTC
Created attachment 847391 [details]
File: environ

Comment 6 collura 2014-01-08 23:32:22 UTC
Created attachment 847392 [details]
File: exploitable

Comment 7 collura 2014-01-08 23:32:24 UTC
Created attachment 847393 [details]
File: limits

Comment 8 collura 2014-01-08 23:32:28 UTC
Created attachment 847394 [details]
File: maps

Comment 9 collura 2014-01-08 23:32:31 UTC
Created attachment 847395 [details]
File: open_fds

Comment 10 collura 2014-01-08 23:32:35 UTC
Created attachment 847396 [details]
File: proc_pid_status

Comment 11 collura 2014-01-08 23:32:38 UTC
Created attachment 847397 [details]
File: var_log_messages

Comment 12 collura 2014-01-09 03:46:25 UTC
no ghostscript updates in koji.fedoraproject.org 
but tried to downgrade and also got this running ghostscript-9.10-4.fc20.x86_64

--- Running report_uReport ---
This problem has already been reported.
https://retrace.fedoraproject.org/faf/reports/254899/
https://bugzilla.redhat.com/show_bug.cgi?id=1050607

Comment 13 Tim Waugh 2014-01-09 13:14:16 UTC
Are you able to reproduce this by trying to print the same job again?

Comment 14 collura 2014-01-11 06:48:36 UTC
yes, i had at least one webpage printjob (dont remember which) that caused the ghostscript crash each time i tried to print it.  

there is some randomness because i always seem to catch it mid-project so havent had time to track it well so not sure of cause so i dont have the link but if i figure out an example of one that causes the issue then i will add it here.


(** i might be tracking multiple bugs in this report so may have to split but hoping same bug because i havent figured what causes the queue clog but both happened at same time in this instance.  

there is a little randomness to the held/held_jobs_printing_after_a_reboot.
'systemctl restart cups.service' didnt seem to effect anything but a reboot usually did.

sometimes a string of jobs seem to print fine then one just stops unepectedly and clogs up the queue and i find out later that i have a queue of unprinted jobs.

sometimes the held job claims that the printer is offline but it can be pinged and seems to be awake if look at its controlpanel and can be printed to from other hosts.)

Comment 15 Tim Waugh 2014-01-13 12:08:50 UTC
Thanks. An example job that triggers it would be very useful.

To aid debugging you can turn on preservation of jobs with "cupsctl PreserveJobFiles=Yes". Then next time a job fails you'll be able to attach the /var/spool/cups/d* file corresponding to the failed job.

I'll leave this as 'needing info' for now.

Comment 16 collura 2014-06-18 08:25:16 UTC
Created attachment 909876 [details]
clog job

had a clog.

i didnt find any '/var/spool/cups/d*' jobs.

there were jobs in /var/spool/cups but they got flushed before i could get a look at them because i had to reboot which successfully restarted printing

i am not sure if it helps but the attached file is one that was in '/var/spool/cups/tmp' and that i had been looking at before the reboot.  
the attached file didnt look remarkable so am wondering if the job before it might have caused the clog but then been cleared before the clog took effect?

will see if i can get another one.

Comment 17 collura 2014-06-18 09:44:32 UTC
regarding comment#16

the clog that happened was characteristic but i just checked abrt and there is no corresponding abrt entry so maybe its a different error and this could be closed?  i havent seen the abrt for some time but the clog continues periodically.  maybe related to my speed of queing a series of print jobs instead of a print job problem?

Comment 18 Tim Waugh 2014-06-18 09:56:50 UTC
What does 'rpm -q ghostscript' say? I think a recent update might fix this.

Before trying an updated ghostscript package, could you please try running this?:

gs -q -dNOPAUSE -dBATCH -dSAFER -sDEVICE=ps2write \
   -sOUTPUTFILE=/dev/null -dLanguageLevel=3 -r1200 \
   -dCompressFonts=false -dNoT3CCITT -dNOINTERPOLATE \
   -c 'save pop' -f cupsS4pXj6-job-that-clogged-que

Do you get any output from that command?

Comment 19 Tim Waugh 2014-06-18 09:58:32 UTC
Also, is there anything in /var/log/cups/error_log from the time you found the clogged job?

Comment 20 collura 2014-06-18 10:20:41 UTC
Created attachment 909937 [details]
snippet of var log messages near time of print que clog during rapid addition of jobs

maybe the original bug issue of abrt is fixed and unrelated to the clog 
because it looks like at a time near this clog 
that avahi was arguing with firefox?  

maybe adding about 1 printjob/second to que on slow machine confused avahi or firefox and i need to open a different bug and/or work slower.

Comment 21 collura 2014-06-18 13:08:17 UTC
Created attachment 910001 [details]
part of the var log cups error_log for comment#19 but also has old gs snipet

Comment 22 collura 2014-06-18 13:24:03 UTC
regarding comment#18

  $ rpm -q ghostscript
     ghostscript-9.14-3.fc20.x86_64

running the 

  "gs -q -dNOPAUSE -dBATCH -dSAFER -sDEVICE=ps2write -sOUTPUTFILE=/dev/null -dLanguageLevel=3 -r1200 -dCompressFonts=false -dNoT3CCITT -dNOINTERPOLATE -c 'save pop' -f cupsS4pXj6-job-that-clogged-que"

produced no output 
and didnt seem to show up in journalctl or /var/log/cups/error_log

Comment 23 Tim Waugh 2014-06-19 08:56:51 UTC
It looks like this is the problem now:

D [14/Jun/2014:05:01:18 -0400] [Job 115] Connecting to <removed-ip6>:631
I [14/Jun/2014:05:01:18 -0400] [Job 115] Connecting to printer.
[...]
D [14/Jun/2014:05:01:18 -0400] [Job 115] Connection error: Invalid argument
E [14/Jun/2014:05:01:18 -0400] [Job 115] The printer is not responding.

Job 115 is for the same destination as (successful) job 114, with device URI dnssd://Lexmark%20MS310dn._ipp._tcp.local/. So it looks like dnssd is deciding on back IPv6 addressses sometimes.

Would it be possible to see what the IPv6 address it tried to connect to was?

Also useful would be the output of this command, as root:

DEVICE_URI=dnssd://Lexmark%20MS310dn._ipp._tcp.local/ \
  /usr/lib/cups/backend/dnssd 1 tim '' 1 '' </dev/null

Comment 24 collura 2014-09-06 11:43:58 UTC
# DEVICE_URI=dnssd://Lexmark%20MS310dn._ipp._tcp.local/ \
>   /usr/lib/cups/backend/dnssd 1 tim '' 1 '' </dev/null
DEBUG: Resolving "Lexmark MS310dn._ipp._tcp.local"...
STATE: +connecting-to-device
DEBUG: Resolving "Lexmark MS310dn", regtype="_ipp._tcp", domain="local."...
DEBUG: Resolved as "ipp://<removed-ipv4-address-of-printer>:631/ipp/print"...
STATE: -connecting-to-device,offline-report
DEBUG: Executing backend "/usr/lib/cups/backend/ipp"...
DEBUG: Sending stdin for job...
DEBUG: update_reasons(attr=0(), s="+connecting-to-device")
DEBUG2: op='+', new_reasons=1, state_reasons=0
STATE: +connecting-to-device
DEBUG: Looking up "<removed-ipv4-address-of-printer>"...
DEBUG: backendWaitLoop(snmp_fd=3, addr=0x7f237fa6cfa8, side_cb=0x7f237f4333b0)

(the backendWaitLoop line seems to be the end of the output, 
only waited about 3minutes after that line though)

Comment 25 Tim Waugh 2014-09-08 10:16:17 UTC
If you run that command several times, do you always get the same output? Does it ever include an IPv6 address?

Comment 26 Tim Waugh 2014-09-08 10:18:42 UTC
(Changing the component and description to describe the IPv6 issue for now.)

Comment 27 Fedora End Of Life 2015-05-29 10:24:07 UTC
This message is a reminder that Fedora 20 is nearing its end of life.
Approximately 4 (four) weeks from now Fedora will stop maintaining
and issuing updates for Fedora 20. It is Fedora's policy to close all
bug reports from releases that are no longer maintained. At that time
this bug will be closed as EOL if it remains open with a Fedora  'version'
of '20'.

Package Maintainer: If you wish for this bug to remain open because you
plan to fix it in a currently maintained version, simply change the 'version' 
to a later Fedora version.

Thank you for reporting this issue and we are sorry that we were not 
able to fix it before Fedora 20 is end of life. If you would still like 
to see this bug fixed and are able to reproduce it against a later version 
of Fedora, you are encouraged  change the 'version' to a later Fedora 
version prior this bug is closed as described in the policy above.

Although we aim to fix as many bugs as possible during every release's 
lifetime, sometimes those efforts are overtaken by events. Often a 
more recent Fedora release includes newer upstream software that fixes 
bugs or makes them obsolete.

Comment 28 Fedora End Of Life 2015-06-29 14:16:57 UTC
Fedora 20 changed to end-of-life (EOL) status on 2015-06-23. Fedora 20 is
no longer maintained, which means that it will not receive any further
security or bug fix updates. As a result we are closing this bug.

If you can reproduce this bug against a currently maintained version of
Fedora please feel free to reopen this bug against that version. If you
are unable to reopen this bug, please file a new report against the
current release. If you experience problems, please add a comment to this
bug.

Thank you for reporting this bug and we are sorry it could not be fixed.


Note You need to log in before you can comment on or make changes to this bug.