From owner-freebsd-hackers  Sun Aug 17 20:06:21 1997
Return-Path: <owner-freebsd-hackers>
Received: (from root@localhost)
          by hub.freebsd.org (8.8.5/8.8.5) id UAA15818
          for hackers-outgoing; Sun, 17 Aug 1997 20:06:21 -0700 (PDT)
Received: from scanner.worldgate.com (scanner.worldgate.com [198.161.84.3])
          by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id UAA15813
          for <hackers@FreeBSD.ORG>; Sun, 17 Aug 1997 20:06:19 -0700 (PDT)
Received: from znep.com (uucp@localhost)
	by scanner.worldgate.com (8.8.5/8.8.5) with UUCP id VAA23890;
	Sun, 17 Aug 1997 21:06:04 -0600 (MDT)
Received: from localhost (marcs@localhost) by alive.znep.com (8.7.5/8.7.3) with SMTP id VAA01395; Sun, 17 Aug 1997 21:05:41 -0600 (MDT)
Date: Sun, 17 Aug 1997 21:05:41 -0600 (MDT)
From: Marc Slemko <marcs@znep.com>
To: The Hermit Hacker <scrappy@hub.org>
cc: hackers@FreeBSD.ORG
Subject: Re: [HACKERS] innd remalloc failure *after* setting MEMDSIZ
In-Reply-To: <Pine.NEB.3.96.970817223912.317S-100000@thelab.hub.org>
Message-ID: <Pine.BSF.3.95.970817210142.27023X-100000@alive.znep.com>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Sender: owner-freebsd-hackers@FreeBSD.ORG
X-Loop: FreeBSD.org
Precedence: bulk

On Sun, 17 Aug 1997, The Hermit Hacker wrote:

> 
> Hi...
> 
> 	After going through "the archives", and finding the suggestion of increasing
> MEMDSIZ to 256Meg, and recompiling/running the new kernel...I'm still getting
> the remalloc 'crash' with innd.
> 
> 	The kernel is 2.2.2-RELEASE, innd is 1.6b3...I have an 'unlimit' at the
> beginning of ~news/etc/rc.news...

What does a ulimit -a run from rc.news after your unlimit say?

It should be MAXDSIZ; was the just a typo, or did you set the wrong thing?

What does the inn log report?

It seems like there may be some VM bogons in 2.2.  I have times when innd
is 40 megs, but 150 megs of swap is being used on the machine; if I flush
a channel that frees 10 megs of memory, I go down to 20 megs of swap in
use.  I haven't looked into it too deeply since I did some changes to the
channel buffer handling to keep the size of them down.  I have had innd
literally eat up 500 megs of VM while still showing as 40 megs in a ps.
This may be related to changes made after 2.2.0.

> 
> 	Can anyone think of something else to try, in order to fix this?
> 
> 	My machine is a P166, with 128Meg of RAM, and 430Meg of SWAP...is datasize
> the only thing that would affect this particular problem?  

Probably.