I received only a few answers to two posts on Solaris 2.1. I got
no `answers' but perhaps rumors and conjectures are worth the bandwidth.
A solaris 2.1 print client when printing to a sun OS 4.1.3 server
Why does lpstat take so long (1/4 hour) to notice that a file has printed?
Don't know. Someone rumored this was a known bug in SVR4. (Can anyone
say for sure?) Someone else reported that lpstat took as long as a
day after the file was printed before lpstat stopped listing it as a
queued print job.
An idle 4.1.3 Sparc generally shows a load average of 0.00. However, an
idle 2.1 IPC generally has a load average near 0.2.
Don't know. Since idle time (reported by top) + top cpu time is
over 99% it isn't one process running 1/5 of the time. Someone
conjectured it is the `biod' replacement. Nfs is multi-threaded,
ever so often all the threads wake up and all could(?) become
ready at the same time. If there are N threads, each taking 1/M
of the time between wakeups yields a load average of N(N+1)/2M but
takes only N/M units of cpu time. If N/M = 0.01 then N > 40 to
get a load average of 0.2. (Better answers?)
There is a way to configure a printer so that banners are optional
(print jobs can request no banner), is there a way to turn them off?
(So banners are never printed.)
A postscript printer seems to use the code in banner.ps
(in /usr/lib/lp/postscript) to print banner pages. Perhaps
replacing this file with
/banner { pop pop pop pop pop } bind def
will turn off the banners on ALL of the postscript printers.
(I haven't tested this yet. But the existing code seems to
require 5 parameters for owner, originator, title, job-id and
printed.)
Questions still with no answers or conjectures:
1. fast filters.
Our current lpr system queries the postscript printers for pages printed
before and after every print job. Is there similar accounting bundled
in 5.1? If not, my understanding is that I will have to add this functionality
to every `fast' filter (i.e. download, postio and postior (what do these last
two do?))
2. allow vs deny.
We currently turn off printing for students who have reached their quota. This
funtionality is handled with a shell script using the `af=shell script' in
the /etc/printcap file. One could use either the allow or the deny feature
to do this in lp. Which (if any) is better? [most printers have most users
allowed, but some printers have only a few allowed.]
This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:07:38 CDT