From: "Austin S. Hemmelgarn" <ahferroin7@gmail.com>
To: Chris Murphy <lists@colorremedies.com>, Liu Bo <bo.li.liu@oracle.com>
Cc: Goffredo Baroncelli <kreijack@inwind.it>,
Btrfs BTRFS <linux-btrfs@vger.kernel.org>
Subject: Re: [PATCH 00/14 RFC] Btrfs: Add journal for raid5/6 writes
Date: Wed, 23 Aug 2017 11:47:17 -0400 [thread overview]
Message-ID: <0236e735-8b68-a0a9-1dc3-e32975b01e29@gmail.com> (raw)
In-Reply-To: <CAJCQCtRrGRgU0u3+qU1pMgsyLfhOWHDrF6CCgVc8c9+JsLuoow@mail.gmail.com>
On 2017-08-23 11:28, Chris Murphy wrote:
> On Wed, Aug 2, 2017 at 2:27 PM, Liu Bo <bo.li.liu@oracle.com> wrote:
>> On Wed, Aug 02, 2017 at 10:41:30PM +0200, Goffredo Baroncelli wrote:
>
>>> What I want to understand, is if it is possible to log only the "partial stripe" RMW cycle.
>>>
>>
>> I think your point is valid if all data is written with datacow. In
>> case of nodatacow, btrfs does overwrite in place, so a full stripe
>> write may pollute on-disk data after unclean shutdown. Checksum can
>> detect errors but repair thru raid5 may not recover the correct data.
>
> What's simpler? raid56 journal for everything (cow, nocow, data,
> metadata), or to apply some limitations to available layouts?
>
> - if raid56, then cow only (no such thing as nodatacow)
This should obviously be something that will be contentious to certain
individuals.
> - permit raid56 for data bg only, system and metadata can be raid1, or raid10
>
> I'm hard pressed thinking of a use case where metadata raid56 is
> beneficial over raid10; a metadata heavy workload is not well suited
> for any parity raid. And if it isn't metadata heavy, then chances are
> you don't even need raid10 but raid1 is sufficient.
Until BTRFS gets n-way replication, raid6 remains the only way to
configure a BTRFS volume to survive more than one device failure.
>
> Of the more complicated ways to solve it:
>
> - journal
> - dynamically sized stripes, so that writes can always be full stripe
> writes, no overwrites, and atomic
> - mixed block groups where only sequential full stripe writes use
> raid56 block group; random and smaller writes go in a raid 1 or 10
> block group.
>
next prev parent reply other threads:[~2017-08-23 15:47 UTC|newest]
Thread overview: 41+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-08-01 16:14 [PATCH 00/14 RFC] Btrfs: Add journal for raid5/6 writes Liu Bo
2017-08-01 16:14 ` [PATCH 01/14] Btrfs: raid56: add raid56 log via add_dev v2 ioctl Liu Bo
2017-08-02 19:25 ` Nikolay Borisov
2017-08-01 16:14 ` [PATCH 02/14] Btrfs: raid56: do not allocate chunk on raid56 log Liu Bo
2017-08-01 16:14 ` [PATCH 03/14] Btrfs: raid56: detect raid56 log on mount Liu Bo
2017-08-01 16:14 ` [PATCH 04/14] Btrfs: raid56: add verbose debug Liu Bo
2017-08-01 16:14 ` [PATCH 05/14] Btrfs: raid56: add stripe log for raid5/6 Liu Bo
2017-08-01 16:14 ` [PATCH 06/14] Btrfs: raid56: add reclaim support Liu Bo
2017-08-01 16:14 ` [PATCH 07/14] Btrfs: raid56: load r5log Liu Bo
2017-08-01 16:14 ` [PATCH 08/14] Btrfs: raid56: log recovery Liu Bo
2017-08-01 16:14 ` [PATCH 09/14] Btrfs: raid56: add readahead for recovery Liu Bo
2017-08-01 16:14 ` [PATCH 10/14] Btrfs: raid56: use the readahead helper to get page Liu Bo
2017-08-01 16:14 ` [PATCH 11/14] Btrfs: raid56: add csum support Liu Bo
2017-08-01 16:14 ` [PATCH 12/14] Btrfs: raid56: fix error handling while adding a log device Liu Bo
2017-08-01 16:14 ` [PATCH 13/14] Btrfs: raid56: initialize raid5/6 log after adding it Liu Bo
2017-08-01 16:14 ` [PATCH 14/14] Btrfs: raid56: maintain IO order on raid5/6 log Liu Bo
2017-08-01 16:14 ` [PATCH 1/2] Btrfs-progs: add option to add raid5/6 log device Liu Bo
2017-08-01 16:14 ` [PATCH 2/2] Btrfs-progs: introduce super_journal_tail to inspect-dump-super Liu Bo
2017-08-01 17:25 ` [PATCH 00/14 RFC] Btrfs: Add journal for raid5/6 writes Roman Mamedov
2017-08-01 17:03 ` Liu Bo
2017-08-01 17:39 ` Austin S. Hemmelgarn
2017-08-01 17:07 ` Liu Bo
2017-08-02 18:47 ` Chris Mason
2018-05-03 19:16 ` Goffredo Baroncelli
2017-08-01 17:28 ` Hugo Mills
2017-08-01 16:56 ` Liu Bo
2017-08-01 18:15 ` Hugo Mills
2017-08-01 17:42 ` Goffredo Baroncelli
2017-08-01 17:24 ` Liu Bo
2017-08-01 22:14 ` Goffredo Baroncelli
2017-08-02 17:57 ` Liu Bo
2017-08-02 20:41 ` Goffredo Baroncelli
2017-08-02 20:27 ` Liu Bo
2017-08-03 4:02 ` Duncan
2017-08-03 4:40 ` Goffredo Baroncelli
2017-08-23 15:28 ` Chris Murphy
2017-08-23 15:47 ` Austin S. Hemmelgarn [this message]
2017-08-25 13:53 ` Goffredo Baroncelli
2017-08-01 21:00 ` Christoph Anton Mitterer
2017-08-01 22:24 ` Goffredo Baroncelli
[not found] <CAL5DHTFA8G5fq=BQ7N7qQimesjygKTiBR2qZF4YBRAsAjB_L5Q@mail.gmail.com>
2019-07-30 16:20 ` Goffredo Baroncelli
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=0236e735-8b68-a0a9-1dc3-e32975b01e29@gmail.com \
--to=ahferroin7@gmail.com \
--cc=bo.li.liu@oracle.com \
--cc=kreijack@inwind.it \
--cc=linux-btrfs@vger.kernel.org \
--cc=lists@colorremedies.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).