From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 8BEB1CA0FE8 for ; Fri, 1 Sep 2023 22:40:05 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1350871AbjIAWkG (ORCPT ); Fri, 1 Sep 2023 18:40:06 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:55018 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1350884AbjIAWkE (ORCPT ); Fri, 1 Sep 2023 18:40:04 -0400 Received: from len.romanrm.net (len.romanrm.net [91.121.86.59]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 7C83A198B for ; Fri, 1 Sep 2023 14:55:49 -0700 (PDT) Received: from nvm (nvm2.home.romanrm.net [IPv6:fd39::4a:3cff:fe57:d6b5]) by len.romanrm.net (Postfix) with SMTP id B2D5040171; Fri, 1 Sep 2023 21:26:37 +0000 (UTC) Date: Sat, 2 Sep 2023 02:26:29 +0500 From: Roman Mamedov To: CoolCold Cc: Linux RAID Subject: Re: raid10, far layout initial sync slow + XFS question Message-ID: <20230902022629.6ab77d48@nvm> In-Reply-To: References: <20230902013700.4c969472@nvm> <20230902020048.356667d4@nvm> X-Mailer: Claws Mail 3.11.1 (GTK+ 2.24.31; x86_64-pc-linux-gnu) MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Precedence: bulk List-ID: X-Mailing-List: linux-raid@vger.kernel.org On Sat, 2 Sep 2023 04:17:46 +0700 CoolCold wrote: > Definitely there is a bottleneck and I very much doubt I'm the first > one facing this - NVMe drives with > 1GB/sec are quite widespread. > > > RAID1 performance at some point if you raise it further to 4096, 8192 or more? > > I can try, for sake of testing, but in terms of practical outcome - > let's imagine with 8MB chunks it reaches maximum - what to do with > that knowledge? Maybe then it would be easier for an mdadm developer to chime in and pinpoint the reason why it might be slow for you at the smaller chunk sizes, provide a hint if there were any commits improving that aspect in kernel versions later than you use, etc. -- With respect, Roman