linux-btrfs.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Anand Jain <anand.jain@oracle.com>
To: linux-btrfs@vger.kernel.org
Cc: dsterba@suse.com, josef@toxicpanda.com
Subject: [PATCH v4 0/3] btrfs: read_policy types latency, device and round-robin
Date: Tue, 19 Jan 2021 23:52:04 -0800	[thread overview]
Message-ID: <cover.1611114341.git.anand.jain@oracle.com> (raw)

v4:
Add rb from Josef in patch 1 and 3.
In patch 1/3, use fs_info instead of device->fs_devices->fs_info.
Drop round-robin policy because my workload (fio random) shows no performance
 gains due to fewer merges at the block layer.

v3:
The block layer commit 0d02129e76ed (block: merge struct block_device and
struct hd_struct) has changed the first argument in the function
part_stat_read_all() in 5.11-rc1. So trickle down its changes in the patch 1/4.

v2:
Fixes as per review comments, as in the individual patches.

rfc->v1:
Drop the tracing patch.
Drop the factor associated with the inflight commands (because there
were too many unnecessary switches).
Few C styles fix.

-----

This patchset adds read policy types latency, device, and round-robin, for the
mirrored raid profiles such as raid1, raid1c3, raid1c4, and raid10. The default
read policy remains as PID, as of now.

Read policy types:
Latency:

Latency policy routes the read IO based on the historical average
wait time experienced by the read IOs on the individual device.

Device:

With the device policy along with the read_preferred flag, you can
set the device for reading manually. Useful to test mirrors in a
deterministic way and helps advance system administrations.

Round-robin (RFC patch):

Alternates striped device in a round-robin loop for reading. To achieve
this first we put the stripes in an array, sort it by devid and pick the
next device.

Test scripts:
=============

I have included a few scripts which were useful for testing.

-------------------8<--------------------------------
Set latency policy on the btrfs mounted at /mnt

Usage example:
  $ readpolicyset /mnt latency

Anand Jain (3):
  btrfs: add read_policy latency
  btrfs: introduce new device-state read_preferred
  btrfs: introduce new read_policy device

 fs/btrfs/sysfs.c   | 57 ++++++++++++++++++++++++++++++++++++++++++-
 fs/btrfs/volumes.c | 60 ++++++++++++++++++++++++++++++++++++++++++++++
 fs/btrfs/volumes.h |  5 ++++
 3 files changed, 121 insertions(+), 1 deletion(-)

-- 
2.28.0


WARNING: multiple messages have this Message-ID (diff)
From: Anand Jain <anand.jain@oracle.com>
To: linux-btrfs@vger.kernel.org
Cc: josef@toxicpanda.com, dsterba@suse.com
Subject: [PATCH v4 0/3, full-cover-letter] btrfs: read_policy types latency, device and round-robin
Date: Wed, 20 Jan 2021 20:34:37 +0800	[thread overview]
Message-ID: <cover.1611114341.git.anand.jain@oracle.com> (raw)
Message-ID: <20210120123437.OVx7ybGaVfmOdZxtpp43qcB_ORHQQs5OzPSzr3ZUGbo@z> (raw)
In-Reply-To: <cover.1611114341.git.anand.jain@oracle.com>

[Only some parts of the cover-letter went through, tying again.].

v4:
Add rb from Josef in patch 1 and 3.
In patch 1/3, use fs_info instead of device->fs_devices->fs_info.
Drop round-robin policy because my workload (fio random) shows no performance
 gains due to fewer merges at the block layer.

v3:
The block layer commit 0d02129e76ed (block: merge struct block_device and
struct hd_struct) has changed the first argument in the function
part_stat_read_all() in 5.11-rc1. So trickle down its changes in the patch 1/4.

v2:
Fixes as per review comments, as in the individual patches.

rfc->v1:
Drop the tracing patch.
Drop the factor associated with the inflight commands (because there
were too many unnecessary switches).
Few C styles fix.

-----

This patchset adds read policy types latency, device, and round-robin, for the
mirrored raid profiles such as raid1, raid1c3, raid1c4, and raid10. The default
read policy remains as PID, as of now.

Read policy types:
Latency:

Latency policy routes the read IO based on the historical average
wait time experienced by the read IOs on the individual device.

Device:

With the device policy along with the read_preferred flag, you can
set the device for reading manually. Useful to test mirrors in a
deterministic way and helps advance system administrations.

Round-robin (RFC patch):

Alternates striped device in a round-robin loop for reading. To achieve
this first we put the stripes in an array, sort it by devid and pick the
next device.

Test scripts:
=============

I have included a few scripts which were useful for testing.

-------------------8<--------------------------------
Set latency policy on the btrfs mounted at /mnt

Usage example:
  $ readpolicyset /mnt latency

Anand Jain (3):
  btrfs: add read_policy latency
  btrfs: introduce new device-state read_preferred
  btrfs: introduce new read_policy device

 fs/btrfs/sysfs.c   | 57 ++++++++++++++++++++++++++++++++++++++++++-
 fs/btrfs/volumes.c | 60 ++++++++++++++++++++++++++++++++++++++++++++++
 fs/btrfs/volumes.h |  5 ++++
 3 files changed, 121 insertions(+), 1 deletion(-)

-- 
2.28.0


             reply	other threads:[~2021-01-20  7:54 UTC|newest]

Thread overview: 15+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-01-20  7:52 Anand Jain [this message]
2021-01-20  7:52 ` [PATCH v4 1/3] btrfs: add read_policy latency Anand Jain
2021-01-20 12:14   ` David Sterba
2021-01-21 10:10     ` Anand Jain
2021-01-21 17:52       ` David Sterba
2021-01-22  8:10         ` Anand Jain
2021-01-30  1:08           ` Anand Jain
2021-02-04 12:30             ` Anand Jain
2021-02-09 21:12               ` Michal Rostecki
2021-02-10  6:14                 ` Anand Jain
2021-01-20  7:52 ` [PATCH v4 2/3] btrfs: introduce new device-state read_preferred Anand Jain
2021-01-21 10:19   ` Anand Jain
2021-01-20  7:52 ` [PATCH v4 3/3] btrfs: introduce new read_policy device Anand Jain
2021-01-20 12:34 ` [PATCH v4 0/3, full-cover-letter] btrfs: read_policy types latency, device and round-robin Anand Jain
2021-01-22  5:52   ` Anand Jain

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=cover.1611114341.git.anand.jain@oracle.com \
    --to=anand.jain@oracle.com \
    --cc=dsterba@suse.com \
    --cc=josef@toxicpanda.com \
    --cc=linux-btrfs@vger.kernel.org \
    /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).