From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <linux-btrfs-owner@vger.kernel.org>
Received: from mail-oi0-f51.google.com ([209.85.218.51]:32838 "EHLO
	mail-oi0-f51.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org
	with ESMTP id S1751231AbcFYVw3 (ORCPT
	<rfc822;linux-btrfs@vger.kernel.org>);
	Sat, 25 Jun 2016 17:52:29 -0400
Received: by mail-oi0-f51.google.com with SMTP id u201so158027625oie.0
        for <linux-btrfs@vger.kernel.org>; Sat, 25 Jun 2016 14:52:29 -0700 (PDT)
MIME-Version: 1.0
In-Reply-To: <CAJCQCtSqO4GNm8kBDuzUXEXYx+54zFgsD6=ARNsRgVUb53LQZw@mail.gmail.com>
References: <CAJCQCtR9uAn58KJKEjCsbyLYJTQVqMx-ghsVp_MjLBF-aiikcg@mail.gmail.com>
 <20160620204049.GA1986@hungrycats.org> <CAJCQCtR5pV53mFyGWxRxm69zwF5_sEvNRRRvOSgnZ1t8KZdc3g@mail.gmail.com>
 <20160621015559.GM15597@hungrycats.org> <CAJCQCtRUUd+moK25N3704ZG54cFrCw1-Uxm2QO-XF9g0=mHazw@mail.gmail.com>
 <20160622203504.GQ15597@hungrycats.org> <5790aea9-0976-1742-7d1b-79dbe44008c3@inwind.it>
 <CAJCQCtRXqSCFZTca+Vwraa0vS-MzLQKEkr=s41Vypc-O0ZDdxQ@mail.gmail.com>
 <20160624014752.GB14667@hungrycats.org> <576CB0DA.6030409@gmail.com>
 <20160624085014.GH3325@carfax.org.uk> <CAA91j0Uqg0FafFTG3NhQt=p8KkYRhTeMU5Bd+JuUxDntP6g8Ng@mail.gmail.com>
 <CAJCQCtQBQAng5_mNJZev+64Z15BSkBkG9f2qmz=ckPRqXRbbWA@mail.gmail.com>
 <576D6C0A.7070502@gmail.com> <CAJCQCtSskA4PC_a8tgQopHFNO83NQ=Gkx406haB7G0nBi5e=2A@mail.gmail.com>
 <c2a320a6-261b-723d-ab83-58f883e6315b@gmail.com> <CAJCQCtSqO4GNm8kBDuzUXEXYx+54zFgsD6=ARNsRgVUb53LQZw@mail.gmail.com>
From: Chris Murphy <lists@colorremedies.com>
Date: Sat, 25 Jun 2016 15:52:26 -0600
Message-ID: <CAJCQCtQhRbL_AWpspudPZzkkGkHT3O9T0QL94Of5LD8agmonwg@mail.gmail.com>
Subject: Re: Adventures in btrfs raid5 disk recovery
To: Chris Murphy <lists@colorremedies.com>
Cc: "Austin S. Hemmelgarn" <ahferroin7@gmail.com>,
        Andrei Borzenkov <arvidjaar@gmail.com>,
        Hugo Mills <hugo@carfax.org.uk>,
        Zygo Blaxell <ce3g8jdj@umail.furryterror.org>, kreijack@inwind.it,
        Roman Mamedov <rm@romanrm.net>,
        Btrfs BTRFS <linux-btrfs@vger.kernel.org>
Content-Type: text/plain; charset=UTF-8
Sender: linux-btrfs-owner@vger.kernel.org
List-ID: <linux-btrfs.vger.kernel.org>

Interestingly enough, so far I'm finding with full stripe writes, i.e.
3x raid5, exactly 128KiB data writes, devid 3 is always parity. This
is raid4. So...I wonder if some of these slow cases end up with a
bunch of stripes that are effectively raid4-like, and have a lot of
parity overwrites, which is where raid4 suffers due to disk
contention.

Totally speculative as the sample size is too small and distinctly non-random.


Chris Murphy