Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 14 Mar 2003 16:05:48 -0500 (EST)
From:      John Baldwin <jhb@FreeBSD.org>
To:        Pawel Jakub Dawidek <nick@garage.freebsd.pl>
Cc:        cvs-all@FreeBSD.org, cvs-src@FreeBSD.org, src-committers@FreeBSD.org, David Schultz <das@FreeBSD.org>, Eivind Eklund <eivind@FreeBSD.org>, Juli Mallett <jmallett@FreeBSD.org>, Poul-Henning Kamp <phk@phk.freebsd.dk>, Garance A Drosihn <drosih@rpi.edu>, Mike Silbersack <silby@silby.com>, "Daniel C. Sobral" <dcs@tcoip.com.br>
Subject:   Re: cvs commit: src/sys/vm ... SIGDANGER
Message-ID:  <XFMail.20030314160548.jhb@FreeBSD.org>
In-Reply-To: <20030314204454.GI567@garage.freebsd.pl>

next in thread | previous in thread | raw e-mail | index | archive | help

On 14-Mar-2003 Pawel Jakub Dawidek wrote:
> On Fri, Mar 14, 2003 at 05:24:42PM -0300, Daniel C. Sobral wrote:
> +> No, there are a couple of other cases:
> +> 
> +> 3) Serious Processing(tm) application blithely allocates memory until 
> +> memory allocation fails. At that point it stops allocating memory. 
> +> Another application then uses some previously untouched allocated 
> +> memory, and first application gets killed losing hours of work.
> +> 
> +> 4) Very Important Server allocates a lot of memory, and gets killed when 
> +> stupid user starts another xterm.
> +> 
> +> Whatever you may or may not think the correct way of handling these two 
> +> cases is, people *have* complained of our present way of handling them.
> 
> Why not choose process to kill by their priority?
> 
> If we got some important processes even without uid=0 we could renice them
> to value less than 0.

I don't think raw priority should be taken into account, but using the
nice value in the algorithm (perhaps as a weight of some sort?) sounds
like a good idea actually.  nice is an existent mechanism for SA's to
mark which processes are more important than others so it seems intuitive
to seek to preserve nice -20 processes at the expense of nice +20
processes.  That also works without changes to existing programs and
will return expected results in current systems w/o the need for major
adjustments.  SIGDANGER, etc. might also be a good idea, but I think
letting nice factor into the equation is a good thing regardless.

-- 

John Baldwin <jhb@FreeBSD.org>  <><  http://www.FreeBSD.org/~jhb/
"Power Users Use the Power to Serve!"  -  http://www.FreeBSD.org/

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe cvs-src" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?XFMail.20030314160548.jhb>