From: Steve Rago Subject: Re: [PATCH] improve the performance of large sequential write NFS workloads Date: Wed, 23 Dec 2009 23:30:14 -0500 Message-ID: <1261629014.13028.160.camel@serenity> References: <1261015420.1947.54.camel@serenity> <1261037877.27920.36.camel@laptop> <20091219122033.GA11360@localhost> <1261232747.1947.194.camel@serenity> <20091222122557.GA604@atrey.karlin.mff.cuni.cz> <1261498815.13028.63.camel@serenity> <20091223183912.GE3159@quack.suse.cz> <1261599385.13028.142.camel@serenity> <1261604952.18047.7.camel@localhost> <1261610013.13028.151.camel@serenity> <1261611898.18047.37.camel@localhost> Mime-Version: 1.0 Content-Type: text/plain Cc: Jan Kara , Wu Fengguang , Peter Zijlstra , "linux-nfs@vger.kernel.org" , "linux-kernel@vger.kernel.org" , "jens.axboe" , Peter Staubach To: Trond Myklebust Return-path: In-Reply-To: <1261611898.18047.37.camel@localhost> Sender: linux-kernel-owner@vger.kernel.org List-ID: On Thu, 2009-12-24 at 00:44 +0100, Trond Myklebust wrote: > > #2 is the difficult one. If you wait for memory pressure, you could > > have waited too long, because depending on the latency of the commit, > > you could run into low-memory situations. Then mayhem ensues, the > > oom-killer gets cranky (if you haven't disabled it), and stuff starts > > failing and/or hanging. So you need to be careful about setting the > > threshold for generating a commit so that the client doesn't run out of > > memory before the server can respond. > > Right, but this is why we have limits on the total number of dirty pages > that can be kept in memory. The NFS unstable writes don't significantly > change that model, they just add an extra step: once all the dirty data > has been transmitted to the server, your COMMIT defines a > synchronisation point after which you know that the data you just sent > is all on disk. Given a reasonable NFS server implementation, it will > already have started the write out of that data, and so hopefully the > COMMIT operation itself will run reasonably quickly. Right. The trick is to do this with the best performance possible. > > Any userland application with basic data integrity requirements will > have the same expectations. It will write out the data and then fsync() > at regular intervals. I've never heard of any expectations from > filesystem and VM designers that applications should be required to > fine-tune the length of those intervals in order to achieve decent > performance. Agreed, except that the more you call fsync(), the more you are stalling the writing, so application designers must use fsync() judiciously. Otherwise they'd just use synchronous writes. (Apologies if I sound like Captain Obvious.) Thanks, Steve > > Cheers > Trond > -- > To unsubscribe from this list: send the line "unsubscribe linux-nfs" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html