[cups.general] Error reading back-channel data: Connection reset by peer

[Michael Sweet]

On Mar 9, 2011, at 2:55 PM, Matt LaPlante wrote:
> We're having problems printing to some Ricoh printers via cups (1.4.5).  The printers will stay online, but files in the associated cups queue printing to them will intermittently get hung with a job (often a PDF) sitting in 'processing' status forever.  The error log seems to indicate a failure writing to the connection.
> ...
> D [09/Mar/2011:17:21:13 +0000] [Job 1612] Error reading back-channel data: Connection reset by peer
> D [09/Mar/2011:17:21:13 +0000] [Job 1612] Set job-printer-state-message to "Unable to write print data: Broken pipe", current level=ERROR
> D [09/Mar/2011:17:21:13 +0000] [Job 1612] Print file sent, waiting for printer to finish...

These messages indicate that the printer is dropping the connection prematurely. There are a bunch of possible reasons for this (including firmware bugs, ACLs not allowing the CUPS system to connect to the printer, an idle timeout that is too short, etc.), so getting a network trace with Wireshark is probably the first thing to look at.

> The job never goes through.  The filters are not failing, and you can find the job, including filter, sitting in the process list.  An strace of the socket://printer-name process shows what appears to be back-channel* traffic over and over again.  The queue gets backed up because the job never stops or goes away until an administrator steps in and cancels it.  I'm open to this being a Ricoh firmware problem, but I'd like to know of I can at least get some insight into what *should* be happening here.  Is the back channel data actually capable of failing a print job, or is it a red herring?  Is it normal for the same values to be repeated over and over again?

The prtMarkSupplies stuff is the SNMP supply/status checking code. You can add "*cupsSNMPSupplies: False" to the PPD file for the printer to disable it if you like.

________________________________________________________________________
Michael Sweet, Senior Printing System Engineer, PWG Chair





-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cups.org/pipermail/cups/attachments/20110309/9113481a/attachment-0001.html>