From mboxrd@z Thu Jan 1 00:00:00 1970 From: Nick Piggin Subject: Re: [patch 00/33] my current vfs scalability patch queue Date: Fri, 4 Sep 2009 09:05:06 +0200 Message-ID: <20090904070506.GA1875@wotan.suse.de> References: <20090904065142.114706411@nick.local0.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: linux-kernel@vger.kernel.org To: linux-fsdevel@vger.kernel.org Return-path: Content-Disposition: inline In-Reply-To: <20090904065142.114706411@nick.local0.net> Sender: linux-kernel-owner@vger.kernel.org List-Id: linux-fsdevel.vger.kernel.org On Fri, Sep 04, 2009 at 04:51:41PM +1000, npiggin@suse.de wrote: > But it is now getting to the point where I will need to get some agreement with > the approach. BTW, yes I'm aware numbers and results are needed. I have some things, but I don't have that big systems to test with, or really interesting workoads. I know google is running into lock contention which is why they proposed the batched iput dput patches. Peter Chubb hit inode lock contention with reaim (don't konw if that is considered realistic). And SGI have had various lock contention problems with NFS fileservers. So aside from reviews and suggestions, what would be most helpful to me would be people willing to test out their problematic workloads with me. It's the only way I can try to get them fixed. Thanks, Nick