From owner-freebsd-current@freebsd.org Sat Sep 8 23:13:25 2018 Return-Path: Delivered-To: freebsd-current@mailman.ysv.freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2610:1c1:1:606c::19:1]) by mailman.ysv.freebsd.org (Postfix) with ESMTP id B2ABD106B87E for ; Sat, 8 Sep 2018 23:13:25 +0000 (UTC) (envelope-from cy.schubert@cschubert.com) Received: from smtp-out-so.shaw.ca (smtp-out-so.shaw.ca [64.59.136.139]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (Client CN "Client", Issuer "CA" (not verified)) by mx1.freebsd.org (Postfix) with ESMTPS id 1DEF37BE6C; Sat, 8 Sep 2018 23:13:24 +0000 (UTC) (envelope-from cy.schubert@cschubert.com) Received: from spqr.komquats.com ([70.67.125.17]) by shaw.ca with ESMTPA id ymPofZv2VwyxUymPpfmxL2; Sat, 08 Sep 2018 17:13:23 -0600 X-Authority-Analysis: v=2.3 cv=NPJhBHyg c=1 sm=1 tr=0 a=VFtTW3WuZNDh6VkGe7fA3g==:117 a=VFtTW3WuZNDh6VkGe7fA3g==:17 a=kj9zAlcOel0A:10 a=JBFolyDoGHsA:10 a=xfDLHkLGAAAA:8 a=YxBL1-UpAAAA:8 a=m1GS6RFLAAAA:8 a=6I5d2MoRAAAA:8 a=MyJtD5Qd0gigxvGbRnoA:9 a=CjuIK1q_8ugA:10 a=UJ0tAi3fqDAA:10 a=IfaqVvZgccqrtc8gcwf2:22 a=Ia-lj3WSrqcvXOmTRaiG:22 a=ZkTy8bY4UdX7Fe4buJt1:22 a=IjZwj45LgO3ly-622nXo:22 Received: from slippy.cwsent.com (slippy [10.1.1.91]) by spqr.komquats.com (Postfix) with ESMTPS id ED236260E; Sat, 8 Sep 2018 16:14:19 -0700 (PDT) Received: from slippy.cwsent.com (localhost [127.0.0.1]) by slippy.cwsent.com (8.15.2/8.15.2) with ESMTP id w88NDJNF059298; Sat, 8 Sep 2018 16:13:19 -0700 (PDT) (envelope-from Cy.Schubert@cschubert.com) Received: from slippy (cy@localhost) by slippy.cwsent.com (8.15.2/8.15.2/Submit) with ESMTP id w88NDH3T059295; Sat, 8 Sep 2018 16:13:17 -0700 (PDT) (envelope-from Cy.Schubert@cschubert.com) Message-Id: <201809082313.w88NDH3T059295@slippy.cwsent.com> X-Authentication-Warning: slippy.cwsent.com: cy owned process doing -bs X-Mailer: exmh version 2.8.0 04/21/2012 with nmh-1.7.1 Reply-to: Cy Schubert From: Cy Schubert X-os: FreeBSD X-Sender: cy@cwsent.com X-URL: http://www.cschubert.com/ To: Matthew Macy cc: Cy Schubert , Jakob Alvermark , Mark Johnston , Subbsd , "allanjude@freebsd.org" , freebsd-current Current Subject: Re: ZFS perfomance regression in FreeBSD 12 APLHA3->ALPHA4 In-Reply-To: Message from Matthew Macy of "Sat, 08 Sep 2018 12:49:31 -0700." Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Sat, 08 Sep 2018 16:13:17 -0700 X-CMAE-Envelope: MS4wfDVGMBIp8/7ZxHqB0awy10bFcPTvR2qU0uriSjmuo3eik06IzUHGBtzP9Tkoqimr1d33H/EV32C/pweU05gA97FkO9fqOmJAVhrTjMiz+dHohx/PDSQH 7FS0sGCKqwqxLKv27nQ6VsoyfdziwDIhwO5EGq+r04WBKwLbdjCiRUUaguwaxdO0Z8RFcq+aITQ+LNwtaoo0dUC7ajRsD2rjUQdfTjl+9qKudXz8QBhVI7Eh ubuibRUC7YOIq0ufcJlVyaK/38QCOwIjFIPy8q35N+k+9zAh/bfDSWdNbrI4rYeMTjE+SbFVn9+wOkawk0AbZA== X-BeenThere: freebsd-current@freebsd.org X-Mailman-Version: 2.1.27 Precedence: list List-Id: Discussions about the use of FreeBSD-current List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sat, 08 Sep 2018 23:13:25 -0000 In message , Matthew Macy writes: > > On Sat, Sep 8, 2018 at 11:03 Cy Schubert wrote: > > > In message , Jakob > > Alvermar > > k writes: > > > > > > I will test the patch below and report back. > > > > Agreed, though IMO your workload and your environment need to be > > understood first. What concerns me about the patch is what impact will > > it have on other workloads. Not evicting data and only metadata could > > impact buildworld -DNO_CLEAN for example. I do a -DNO_CLEAN > > buildworlds, sometimes -DWORLDFAST. Adjusting vfs.zfs.arc_meta_limit to > > the same value as vfs.zfs.arc_max improved my buildworld/installworld > > performance. In addition disabling atime for the ZFS dataset containing > > /usr/obj also improved buildworld/installworld performance by reducing > > unnecessary (IMO) metadata writes. I think evicting metadata only might > > cause a new set of problems for different workloads. (Maybe this should > > be a sysctl?) > > > > Mark's suggested change would just restore the default behavior from before > balanced pruning was imported. Balanced pruning is dependent on code that > didn't exist in FreeBSD - unfortunately when I did the import I did not > pull in all the dependencies. Mid freeze, the least disruptive path is to > disable the new behavior. Post branch we can restore it as a default, > possibly contingent on the amount of memory in the system. There is > precedent for this with prefetch. Ok. -- Cheers, Cy Schubert FreeBSD UNIX: Web: http://www.FreeBSD.org The need of the many outweighs the greed of the few.