From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-wm0-f48.google.com ([74.125.82.48]:37506 "EHLO mail-wm0-f48.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1758950AbcJ1KMD (ORCPT ); Fri, 28 Oct 2016 06:12:03 -0400 Received: by mail-wm0-f48.google.com with SMTP id 140so68392683wmv.0 for ; Fri, 28 Oct 2016 03:12:02 -0700 (PDT) Subject: Re: discard improvements References: <1476735753-5861-1-git-send-email-hch@lst.de> From: Avi Kivity Message-ID: <64a1c5dc-4a36-1ef6-606e-b382d36161ef@scylladb.com> Date: Fri, 28 Oct 2016 13:11:55 +0300 MIME-Version: 1.0 In-Reply-To: <1476735753-5861-1-git-send-email-hch@lst.de> Content-Type: text/plain; charset=windows-1252; format=flowed Content-Transfer-Encoding: 7bit Sender: linux-xfs-owner@vger.kernel.org List-ID: List-Id: xfs To: Christoph Hellwig , linux-xfs@vger.kernel.org Cc: michaelcallahan@fb.com On 10/17/2016 11:22 PM, Christoph Hellwig wrote: > iThis series improves XFS discard support so that we don't stall the > log commit thread and improve performance for workloads with lots > of deletions. > > The first patch is a fix for a long-standing allocator issue when we > have busy extents in the way of an allocation. Without the async > discard support it was harder to hit, but it's basically been there > since we added busy extent tracking and non-synchronous delete > transactions. > > The second patch is the main async discard support, and the third > one allows to batch up multiple extents into a single discard bio, > which is a common occurance. > With these patches, is it reasonable to run with discard enabled, or is more work needed?