Path: utzoo!utgpu!jarvis.csri.toronto.edu!rutgers!cbmvax!grr
From: grr@cbmvax.UUCP (George Robbins)
Newsgroups: comp.unix.ultrix
Subject: Re: excessive "init" forking
Message-ID: <7438@cbmvax.UUCP>
Date: 25 Jul 89 18:05:18 GMT
References: <8490@batcomputer.tn.cornell.edu>
Reply-To: grr@cbmvax.UUCP (George Robbins)
Distribution: na
Organization: Commodore Technology, West Chester, PA
Lines: 27

In article <8490@batcomputer.tn.cornell.edu> eric@geology.tn.cornell.edu writes:
> We have a nice new DECstation 3100 running Ultrix 3.0 and DECwindows.
> Unfortunately it is running so slowly as to be essentially useless...
> 		.... It appears that "/etc/init" is constantly forking
> off copies of itself, at a rate of about 50-100 times a minute. Each process
> lasts only a short time but meanwhile uses several hundred K of memory. At
> any one time "ps -aux" shows 5-10 "/etc/init -a" processes running.

Sounds pretty perverse.  If init is constantly forking, it would seem that
it's constantly trying to set something up that is failing.  Maybe look
in syslog for clues, you don't mention anything coming out on the console...

Is /etc/ttys reasonable?  You're not trying to spawn off stuff on serial/
psuedo/LAT devices that aren't defined in /dev or are otherwise defective?

Any resource problems?  Swap space or virtual memory size?  Does this
happen from moment 1, or only take off after some random delay/event?

Of course this could be something intrinsic to the operation of DECwindows,
something of which I know nothing, but suspect uses some hooks into init
since there are already funny "window manager" hooks in BSD init before
DEC got to it.

-- 
George Robbins - now working for,	uucp: {uunet|pyramid|rutgers}!cbmvax!grr
but no way officially representing	arpa: cbmvax!grr@uunet.uu.net
Commodore, Engineering Department	fone: 215-431-9255 (only by moonlite)