From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753408Ab2ITHbM (ORCPT ); Thu, 20 Sep 2012 03:31:12 -0400 Received: from tee.schottelius.org ([77.109.138.222]:59211 "EHLO mx3.schottelius.org" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1753139Ab2ITHbK (ORCPT ); Thu, 20 Sep 2012 03:31:10 -0400 Date: Thu, 20 Sep 2012 09:32:45 +0200 From: Nico Schottelius To: David Rientjes Cc: Nico Schottelius , linux-kernel@vger.kernel.org Subject: Re: Out of memory on 3.5 kernels Message-ID: <20120920073245.GA1327@schottelius.org> Mail-Followup-To: Nico Schottelius , David Rientjes , linux-kernel@vger.kernel.org References: <20120920060202.GB10758@schottelius.org> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: User-Agent: echo $message | gpg -e $sender -s | netcat mailhost 25 X-Netzseite: http://www.nico.schottelius.org/ X-System-Info: brief (Linux 3.5.4-1-ARCH x86_64) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Hey David, David Rientjes [Wed, Sep 19, 2012 at 11:54:54PM -0700]: > On Thu, 20 Sep 2012, Nico Schottelius wrote: > > > for some time the slab usage is quite high (~3.2 GiB) and whenever I try > > to do a backup with ccollect (which is using rsync), a lot of processes > > (like chromium) are being killed over night. > > > > Yeah, over 81% of your memory is consumed by slab and that doesn't shrink > over the duration of the log you posted. It would be interesting to see > if there was a leak somewhere: try looking at slabtop and determining if > you have a cache using an egregious amount of memory; I already had to reboot, because no now process was starting / only starting extremly slowly. > if there aren't any > clear winners, then we'll still see what is hogging most of your memory > and see how it compares to 3.4.2. You can also check if there is a leak > using kmemleak, but this requires a reboot (see > Documentation/kmemleak.txt). I'll have a look at this one after the next reboot; I also have to update to a recent kernel and enable CONFIG_DEBUG_KMEMLEAK, which is missing in Archlinux' kernel. Thanks for the pointers, it can only take a day until I have this problem again - will then follow up this mail. Cheers, Nico -- PGP key: 7ED9 F7D3 6B10 81D7 0EC5 5C09 D7DC C8E4 3187 7DF0