Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!uunet!husc6!cmcl2!brl-adm!adm!bzs@bu-cs.bu.EDU
From: bzs@bu-cs.bu.EDU (Barry Shein)
Newsgroups: comp.unix.wizards
Subject: How to set ulimit in SysV
Message-ID: <9807@brl-adm.ARPA>
Date: Thu, 15-Oct-87 19:47:34 EDT
Article-I.D.: brl-adm.9807
Posted: Thu Oct 15 19:47:34 1987
Date-Received: Sat, 17-Oct-87 09:53:41 EDT
Sender: news@brl-adm.ARPA
Lines: 35


>If you want to raise the ulimit for
>login sessions, GET THE SOURCE to /bin/login and put the ulimit call
>in there before it setuid's and execs the shell.
...
>  "terribly broken and a major oversight"????  Sorry, I bet they did not
>know that you had an INFINITE amount of space available and were not
>concerned with limiting it.  Since I bet somehow that you do not have
>an infinite space, you probably DO need a limiter.  You setting it to
>-1 proves the fact that you are not infallible and might screw up again!!!

You must have worked for the VMS group in a previous life (it can't
be wrong, that's the way we implemented it!)

What's wrong and terribly broken is not that the feature exists, it's
that you can't set it to a value reasonable for your site without the
source (or various wizardry.) All by itself it can cripple a system's
useability.

Actually, ulimit is completely broken anyhow, see BSD's disk quota
system for the kind of thing adults implement. This is not a BSD v
SYSV thing, a disk quota system would fit just fine into SYSV without
interfering with their precious SVID. We can't even consider SYSV
without a rational disk quota system (would you like to manage a
15,000 user system, mostly students, without something better than
ulimit and the hacks you suggest above?) I believe SYSV is about the
only O/S people are actually selling for money these days without some
sort of disk quota system, I may be wrong, but be careful who you name
as your equivalents. How hard could it be to manage four integers? It
boggles the mind.

I think you're acting as an unneeded apologist for an obvious
deficiency.

	-Barry Shein, Boston University