CUPS, printcodes, stale UTMP entries, and terminal hangs

[Jeremy Anderson]

Good afternoon, all.

I have a couple of questions that are stumping me.
My environment is thus:
filePro 5.0.13D4 running on SuSE 9.1 x86.

Most of my users connect via SSH, using Putty running a variety of 
Windows OSs.  Some users use dumb terminals attached to a Digi 
portserver.  My printers are a mix of serial-attached (again, via the 
Digi), laser printers with built-in network capability, and printers 
which are using TCP/IP print servers with lpd functionality.  All 
printing is done through CUPS.

That said, I notice several peculiar things.

#1)  Some of my okidata ml320 printers are attached via the Digi.  Two 
others have TCP/IP print servers.  Before I upgraded, everything was 
hunky dory.  Now, however, the first print job shows 5i4i  in the upper 
left hand corner of the first item of each print job.  I suspect that 
CUPS is not handling the print codes properly.  Is anyone else using 
oki320s through CUPS?  I note that in the print job in question, I'm 
sending through a print code of 4, which shows as "reset printer".

#2)  Stale UTMP entries:  My users are all exec'ing /usr/local/bin/p 
from their login scripts (actually, through /etc/profile).  When they 
log out, their entries are not cleaned up from UTMP.  e.g., they still 
show up in the output from the "who" and "w" command, though they have 
no active processes in a ps listing.    I do not know if this is a SuSE, 
OpenSSH or FilePro issue.  I mention it because I have a shell script 
and small utility which will clean out the excess entries, in case 
anyone wants it.  I also was curious if anyone else is seeing this behavior.

#3)  Some of my users are reporting hung sessions, particularly when 
backing out of screens.  That is, when they are attempting to return to 
the menu.  I have noticed it myself, generally when leaving *cabe or 
dmoedef.  I have not been able to isolate a particular sequence of 
actions which will guarantee a terminal hang.  Has anyone else seen this 
under Linux?  It affects users connecting via TCP/IP w/SSH, and dumb 
terminal users (serial -> Digi -> telnet) as well.  We have also seen 
this happen on the local console.  This leads me to suspect it is not a 
network issue.  Any tips or suggestions?

Thanks for any help you can lend!

Jeremy
-- 
Jeremy Anderson			      jeremy (at) plunketts.net
IT Manager, Plunkett's Pest Control   Author, Multitool Linux