From owner-freebsd-geom@freebsd.org Fri Jul 3 20:18:23 2020 Return-Path: Delivered-To: freebsd-geom@mailman.nyi.freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2610:1c1:1:606c::19:1]) by mailman.nyi.freebsd.org (Postfix) with ESMTP id 9AAA53582E6 for ; Fri, 3 Jul 2020 20:18:23 +0000 (UTC) (envelope-from pawel@dawidek.net) Received: from mail.dawidek.net (garage.dawidek.net [91.121.88.72]) by mx1.freebsd.org (Postfix) with ESMTP id 49z5rH2Cb8z3SHS; Fri, 3 Jul 2020 20:18:22 +0000 (UTC) (envelope-from pawel@dawidek.net) Received: from [192.168.250.156] (c-67-169-9-63.hsd1.ca.comcast.net [67.169.9.63]) by mail.dawidek.net (Postfix) with ESMTPSA id 22FE8445; Fri, 3 Jul 2020 22:18:16 +0200 (CEST) Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable Mime-Version: 1.0 (1.0) Subject: Re: Single-threaded bottleneck in geli From: =?utf-8?Q?Pawe=C5=82_Jakub_Dawidek?= In-Reply-To: Date: Fri, 3 Jul 2020 13:18:13 -0700 Cc: freebsd-geom@freebsd.org Message-Id: <48AB9191-5984-4BE0-BF98-7548424D06BF@dawidek.net> References: To: Alan Somers X-Mailer: iPhone Mail (17F80) X-Rspamd-Queue-Id: 49z5rH2Cb8z3SHS X-Spamd-Bar: ---- Authentication-Results: mx1.freebsd.org; none X-Spamd-Result: default: False [-4.00 / 15.00]; REPLY(-4.00)[]; ASN(0.00)[asn:16276, ipnet:91.121.0.0/16, country:FR] X-BeenThere: freebsd-geom@freebsd.org X-Mailman-Version: 2.1.33 Precedence: list List-Id: GEOM-specific discussions and implementations List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Fri, 03 Jul 2020 20:18:23 -0000 Hi Alan, why do you think it will hurt single-threaded performance? --=20 Pawe=C5=82 Jakub Dawidek > On Jul 3, 2020, at 12:30, Alan Somers wrote: >=20 > =EF=BB=BFI'm using geli, gmultipath, and ZFS on a large system, with hundr= eds of > drives. What I'm seeing is that under at least some workloads, the overal= l > performance is limited by the single geom kernel process. procstat and > kgdb aren't much help in telling exactly why this process is using so much= > CPU, but it certainly must be related to the fact that over 15,000 IOPs ar= e > going through that thread. What can I do to improve this situation? Woul= d > it make sense to enable direct dispatch for geli? That would hurt > single-threaded performance, but probably improve performance for highly > multithreaded workloads like mine. >=20 > Example top output: > PID USERNAME PRI NICE SIZE RES STATE C TIME WCPU COMMAND > 13 root -8 - 0B 96K CPU46 46 82.7H 70.54% > geom{g_down} > 13 root -8 - 0B 96K - 9 35.5H 25.32% > geom{g_up} >=20 > -Alan > _______________________________________________ > freebsd-geom@freebsd.org mailing list > https://lists.freebsd.org/mailman/listinfo/freebsd-geom > To unsubscribe, send any mail to "freebsd-geom-unsubscribe@freebsd.org"