Return-Path: linux-nfs-owner@vger.kernel.org Received: from relay3.sgi.com ([192.48.152.1]:33157 "EHLO relay.sgi.com" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1752811Ab3AYVNR (ORCPT ); Fri, 25 Jan 2013 16:13:17 -0500 Date: Fri, 25 Jan 2013 15:13:16 -0600 From: Ben Myers To: "J. Bruce Fields" Cc: Olga Kornievskaia , linux-nfs@vger.kernel.org, Jim Rees Subject: Re: sunrpc: socket buffer size tuneable Message-ID: <20130125211316.GX30652@sgi.com> References: <20130125192935.GA32470@sgi.com> <20130125202107.GD29596@fieldses.org> <20130125203521.GE29596@fieldses.org> <20130125205145.GF29596@fieldses.org> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii In-Reply-To: <20130125205145.GF29596@fieldses.org> Sender: linux-nfs-owner@vger.kernel.org List-ID: Hey Bruce, On Fri, Jan 25, 2013 at 03:51:45PM -0500, J. Bruce Fields wrote: > On Fri, Jan 25, 2013 at 03:35:21PM -0500, J. Bruce Fields wrote: > > On Fri, Jan 25, 2013 at 03:21:07PM -0500, J. Bruce Fields wrote: > > > > The minimal overflow fix did not resolve the timeouts. > > > > > > OK, thanks, that's expected. > > > > > > > I will test with this to see if it resolves the timeouts: > > > > > > And I'd expect that to do the job--but at the expense of some tcp > > > bandwidth. So you end up needing your other module parameters to get > > > the performance back. > > > > Also, what do you see happening on the server in the problem case--are > > threads blocking in svc_send, or are they dropping replies? I'm working on getting a dump to determine this. Maybe there is an easier way? > And what exactly is your test? Boot of a 144 node nfs root cluster. The timeouts happen when loading a ~14mb kernel module. Thanks, Ben