Hi!
Well, there have been plenty of printer questions lately, and here is another
one, perhaps reminding of the other ones.
We have about 100 HP4 and HP3 laserjets connected to our 3000/600, OSF/1 3.0,
via xyplex terminal servers running a xyplex implementation of the lpd daemon.
The terminal servers are xyplex 1600 and 1640 running version 5.3.1, and the
printers are of course serial attached to the servers.
This is a typical printcap entry:
#ABC10
10lp10abc10:
:lf=/usr/adm/lp10err:
:lp=:
:rm=X37.svt.se:
:rp=10:
:sd=/usr/spool/lpd10:
The problem is that sometimes the printer queues hang in the alpha. To restart
them, lpc abort, lpc clean, lpc start is required. I cannot however reproduce
the problem. This reminds of a LAN Manager problem Thomas Meyer
(meyert_at_dla-marbach.de) reported in january.
We also get a lot of garbage in the log files, which several other reported.
Jun 20 17:46:43 kayo lpd[4418]: pid=6835 (sig=-1,err=0)
Jun 20 17:47:05 kayo lpd[4418]: pid=9444 (sig=-1,err=0)
These entries occur each time a job is printed.
We have had a long discussion with DEC Sweden over this, and change lpd and lpr
to later versions.
I really don't know how to attack this. Is the problem somewhere within the lpd
system in the alpha, since restarting the queues never includes the xyplexes?
Or are we facing a bad implementation of lpd in the xyplexes? Maybe we should
attack it from the strange log entries?! (But we had those before using lpd,
when we had a xyplex supplied filter in the alpha to print on terminal servers.
The filter didn't work very well over our WAN however).
Any ideas are very welcome!
Regards,
Johan Larsson
Swedish Television
Stockholm SWEDEN
Received on Tue Jun 20 1995 - 18:37:28 NZST