This page lists known issues or bugs with PaperCut on Novell OES Linux.
Are you going to Brainshare? Of so, come and visit PaperCut at booth 1012!
Rare crash bug in iPrint (ipsmd) on OES2 Linux SP1
The default iPrint that ships with OES2 Linux SP1 contains a bug where canceling a job may occasionally crash ipsmd. This is a bug in iPrint, however the introduction of PaperCut may "exaggerated" it. On a normal network, the "cancel job" operation would only be used occasionally - say by an administrator cancelling a job to unblock a jammed print queue. PaperCut being a print control application however cancels jobs from time to time as part of standard operation - e.g. a user running out of quota. The fact that the cancel operation is called more often causes this bug to surface more often.
Symptoms:
- Errors may appear in PaperCut's print-provider.log - Unable to cancel job error number: 1349320716 (0x506D0012)
- Errors may appear in the iPrint logs (ipsmd.log)
- The
papercut accounting autoload command may be listed as a <defunct> process (this is because the parent process has died and left us in a zombie/orphaned state as explained here).
Known Workarounds:
- Contact Novell support and ask for the latest iprint update compatible with PaperCut. It's filed under TID 7004872
Crash bug in iPrint (ipsmd) on OES2 Linux SP2
The default iPrint that ships with OES2 Linux SP2 contains a bug where after enabling PaperCut via the Accounting autoload command iPrint may crash after receiving a print job.
Symptoms:
- Jobs do not print after enabling PaperCut on the printer
- Errors may appear in the iPrint logs (ipsmd.log)
Known Workarounds:
- Contact Novell support and ask for the latest iprint update compatible with PaperCut. It's filed under TID 7004872. Ensure the update is suitable for OES2 SP2 and your architecture. The latest version we've stress tested with success is 6.2.20091222 .
General Discussion/Comments (History)
- Update 14th April 2009 - The Novell iPrint team have released a patch, however it is yet to make it out via official channels. Please email support@papercut.com for more information and details on how to access the "unofficial" patch!
- Update 17th June 2009 - The Novell iPrint team have again supplied us with a new set of patches. Again we are able to supply on request. Please email support@papercut.com for more information and details on how to access patch!
- Update 23rd June 2009 - The next release of PaperCut (version 9.5) will have a number of changes that we hope will reduce the probability of the iPrint bug occurring (e.g. thread synchronization changes). The following hotfix may be applied to existing 9.x installs:
su - papercut
cd ~
tar -zcf pre-fix-backup.tgz providers/print/linux-*
wget http://www.papercut.com/anonftp/pub/pcng/previews/novell-oes-hotfix-2009-06-24.tgz
tar -zxf novell-oes-hotfix-2009-06-24.tgz
su - root
/home/papercut/providers/print/linux-x64/roottasks
Then restart The iPrint Print Manager or simply restart the server to apply.
- Update 4th August 2009 - We've been working with the iPrint team to QA a recent iPrint build (20090731-1 begin_of_the_skype_highlighting 20090731-1 end_of_the_skype_highlighting begin_of_the_skype_highlighting 20090731-1 end_of_the_skype_highlighting). This version shows lots of promise. We've run thousands of jobs through via our automated testing procedures and have not had any crashes. We're working with Novell to make this an official release that we can make available to PaperCut users.
- Update 14th August 2009 - Novell have decided to make the 20090731-1 avaliable as an official patch. That's the version we've tested in house with great promise. Should be avaliable in the next few days. I'll post an update here.
- Update 11th September 2009 - Hooray! Looks like we've finally found and fixed the remaining race-condition related bugs. We've put some additional synchronization logic in our code and also the iPrint team has done the same. We were seeing some issues on customer sites running on some very fast multi-processor systems and the latest round of fixes has addressed these issues. Novell are going to make the update avaliable as an official update/hotfix soon. If you're running on a fast multi-core/processor system and need access to the fix now, please email us at support.
- Update October 2009 - Latest 9.6 release of PaperCut with latest iPrint patches have proved to be very stable with no issues reported on the high volume sites. We've also now got PaperCut running on clustered iPrint servers. Official cluster setup documentation will be incorporated into the 9.7 release. If you require early access to cluster support, please email us at support.
- Update October 2009 - We've had a few of the larger sites sites report slow printing performance after applying one of the Novell patches provided to address stability issues above. If you're seeing slow printing, please email support and we can supply a newer patch that addresses this. More information here under TID 7004872.
- Update 28th October 2009 - PaperCut now works on OES Linux clusters. 9.7 will include official setup documentation however if you require early access, please email us at support.
- Update November 2009 - There seems to be a regression OES2 SP2. After upgrading, some customers have reported that the iPrint server crashes. We're working with Novell to resolve this issue. In the meantime we recommend sticking to SP1 and delaying any upgrade.
- Update December 2009 - Novell have issued a hotfix for the regression in OES2 SP2 (listed as bug 555918). Customer's running SP2 should ensure they are running version 6.2.20091222 or higher. We have this patch avaliable, however in line with Novell's support procedure, please contact Novell to obtain this patch through official channels.
- 15th January 2010 - Novell have issued a hotfix for the regression in OES2 SP2 (listed as bug 555918). It's now avaliable under official support channels via TID 7004872.
The PaperCut development team will continue to post updates here as they become available.
keywords: zombie process, oes linux, novell iprint, crash dump