Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!seismo!umcp-cs!chris
From: chris@umcp-cs.UUCP (Chris Torek)
Newsgroups: net.unix-wizards
Subject: Re: timeout table overflow
Message-ID: <3399@umcp-cs.UUCP>
Date: Sat, 13-Sep-86 17:40:23 EDT
Article-I.D.: umcp-cs.3399
Posted: Sat Sep 13 17:40:23 1986
Date-Received: Sat, 13-Sep-86 21:32:15 EDT
References: <1010@batcomputer.TN.CORNELL.EDU>
Reply-To: chris@umcp-cs.UUCP (Chris Torek)
Organization: University of Maryland, Dept. of Computer Sci.
Lines: 36

In article <1010@batcomputer.TN.CORNELL.EDU> hurf@batcomputer.UUCP
(Hurf Sheldon) writes:
[Vax crashes with "panic: timeout table overflow"]
>how can I increase the table size and is there a 
>way already devised for flushing the table periodically?

You do not want to flush the table.  The things in it are important!

The table size is determined by `ncallout'.  This is a `configurable'
parameter, appearing in /sys/YOUR_MACHINE/param.c.  This file is
normally copied from /sys/conf/param.c, which says

	int ncallout = 16 + NPROC;

where NPROC is `#define'd to something depending on MAXUSERS, where
MAXUSERS comes from your config file in /sys/conf/YOUR_MACHINE.
If you are running out of timeout table entries, you probably have
maxusers set too low too; but you can, if you wish, just raise
ncallout, either by editing the private param.c, or (more quickly)
with adb:

	% adb -w /vmunix
	ncallout?D
	_ncallout:
	_ncallout:	122
	ncallout?W 0t250
	_ncallout:
	_ncallout:	7a		=	fa
	$q
	%

This would raise ncallout from (e.g.) 122 decimal to 250 decimal.
-- 
In-Real-Life: Chris Torek, Univ of MD Comp Sci Dept (+1 301 454 1516)
UUCP:	seismo!umcp-cs!chris
CSNet:	chris@umcp-cs		ARPA:	chris@mimsy.umd.edu