[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <alpine.DEB.2.02.1311251730160.27270@chino.kir.corp.google.com>
Date: Mon, 25 Nov 2013 17:31:28 -0800 (PST)
From: David Rientjes <rientjes@...gle.com>
To: Jörn Engel <joern@...fs.org>
cc: Michal Hocko <mhocko@...e.cz>, linux-mm@...ck.org,
Greg Thelen <gthelen@...gle.com>,
Glauber Costa <glommer@...il.com>,
Mel Gorman <mgorman@...e.de>,
Andrew Morton <akpm@...ux-foundation.org>,
Johannes Weiner <hannes@...xchg.org>,
KOSAKI Motohiro <kosaki.motohiro@...fujitsu.com>,
Rik van Riel <riel@...hat.com>,
Hugh Dickins <hughd@...gle.com>,
LKML <linux-kernel@...r.kernel.org>
Subject: Re: user defined OOM policies
On Thu, 21 Nov 2013, Jörn Engel wrote:
> One ancient option I sometime miss was this:
> - Kill the biggest process.
>
That's what both the system and memcg oom killer currently do unless you
change how "biggest process" is defined with /proc/pid/oom_score_adj. The
goal is the kill a single process with the highest rss to avoid having to
kill many different processes.
Powered by blists - more mailing lists