[cups] TCP window size gets set to zero and printing times out/stops

[Michael Sweet]

Ed,

A Wireshark trace might be useful here, but I'd start with Red Hat to see if there is a TCP setting you need to tweak for your network (or something that is a bug).


> On Jul 27, 2015, at 3:17 PM, Smith, Ed <ed.smith at daikinapplied.com> wrote:
> 
> Greetings,
> 
> we've been having a problem with print jobs stopping, so we changed the error-policy to retry. Now print restarts,
> often to completion on the 2nd attempt (but not always). The problem can surface on new and old printers alike, short and
> long reports, running raw, with reports preformatted by server's apps as Postscript, and postscript option installed and running on printers .
> We started experiencing this issue when we migrated from solaris 10/previous cups version (1.2.6, I believe) to RHEL Linux 5.9\cups 1.3.7. The older environment ran on our same network just fine, I don't recall ever having this issue then, it ran for years OK this way. My wireshark sniffs suggests the TCP window size gets set to 0, and communication between server and printer
> just stops. Old and new printers, random, so I don't think it's the firmware. Any clue as to what we might look at next?  Thanks!  -Ed Smith
> 
> Thanks.
> 
> Ed J Smith
> 
> 
> _______________________________________________
> cups mailing list
> cups at cups.org
> https://www.cups.org/mailman/listinfo/cups

_________________________________________________________
Michael Sweet, Senior Printing System Engineer, PWG Chair