public inbox for linux-btrfs@vger.kernel.org
 help / color / mirror / Atom feed
From: Anand Jain <anand.jain@oracle.com>
To: Christoph Hellwig <hch@lst.de>
Cc: Chris Mason <clm@fb.com>, Josef Bacik <josef@toxicpanda.com>,
	David Sterba <dsterba@suse.com>,
	Damien Le Moal <damien.lemoal@wdc.com>,
	Naohiro Aota <naohiro.aota@wdc.com>,
	Johannes Thumshirn <johannes.thumshirn@wdc.com>,
	Qu Wenruo <wqu@suse.com>, Jens Axboe <axboe@kernel.dk>,
	"Darrick J. Wong" <djwong@kernel.org>,
	linux-block@vger.kernel.org, linux-btrfs@vger.kernel.org,
	linux-fsdevel@vger.kernel.org
Subject: Re: [PATCH 04/34] btrfs: remove the direct I/O read checksum lookup optimization
Date: Wed, 25 Jan 2023 15:42:21 +0800	[thread overview]
Message-ID: <8d3e27aa-7b65-e08d-7dc1-a8db2d3f1389@oracle.com> (raw)
In-Reply-To: <20230124195531.GA16743@lst.de>



On 25/01/2023 03:55, Christoph Hellwig wrote:
> On Tue, Jan 24, 2023 at 10:55:25PM +0800, Anand Jain wrote:
>>
>> I was curious about the commit message.
>> I ran fio to test the performance before and after the change.
>> The results were similar.
>>
>> fio --group_reporting=1 --directory /mnt/test --name dioread --direct=1
>> --size=1g --rw=read  --runtime=60 --iodepth=1024 --nrfiles=16 --numjobs=16
>>
>> before this patch
>> READ: bw=8208KiB/s (8405kB/s), 8208KiB/s-8208KiB/s (8405kB/s-8405kB/s),
>> io=481MiB (504MB), run=60017-60017msec
>>
>> after this patch
>> READ: bw=8353KiB/s (8554kB/s), 8353KiB/s-8353KiB/s (8554kB/s-8554kB/s),
>> io=490MiB (513MB), run=60013-60013msec
> 
> That's 4k reads.  The will benefit from the inline csum array in the
> btrfs_bio, but won't benefit from the existing batching, so this is
> kind of expected.
> 
> The good news is that the final series will still use the inline
> csum array for small reads, while also only doing a single csum tree
> lookup for larger reads, so you'll get the best of both worlds.
> 

Ok. Got this results for the whole series from an aarch64 
(pagesize=64k); Results finds little improvement/same.


Before:
Last commit:
b3b1ba7b8c0d btrfs: skip backref walking during fiemap if we know the 
leaf is shared

---- mkfs.btrfs /dev/vdb ..... :0 ----
---- mount -o max_inline=0 /dev/vdb /btrfs ..... :0 ----
---- fio --group_reporting=1 --directory /btrfs --name dioread 
--direct=1 --size=1g --rw=read --runtime=60 --iodepth=1024 --nrfiles=16 
--numjobs=16 | egrep "fio|READ" ..... :0 ----
fio-3.19
    READ: bw=6052MiB/s (6346MB/s), 6052MiB/s-6052MiB/s 
(6346MB/s-6346MB/s), io=16.0GiB (17.2GB), run=2707-2707msec

---- mkfs.btrfs /dev/vdb ..... :0 ----
---- mount -o max_inline=64K /dev/vdb /btrfs ..... :0 ----
---- fio --group_reporting=1 --directory /btrfs --name dioread 
--direct=1 --size=1g --rw=read --runtime=60 --iodepth=1024 --nrfiles=16 
--numjobs=16 | egrep "fio|READ" ..... :0 ----
fio-3.19
    READ: bw=6139MiB/s (6437MB/s), 6139MiB/s-6139MiB/s 
(6437MB/s-6437MB/s), io=16.0GiB (17.2GB), run=2669-2669msec


After:
last commit:
b488ab9aed15 iomap: remove IOMAP_F_ZONE_APPEND

---- mkfs.btrfs /dev/vdb ..... :0 ----
---- mount -o max_inline=0 /dev/vdb /btrfs ..... :0 ----
---- fio --group_reporting=1 --directory /btrfs --name dioread 
--direct=1 --size=1g --rw=read --runtime=60 --iodepth=1024 --nrfiles=16 
--numjobs=16 | egrep "fio|READ" ..... :0 ----
fio-3.19
    READ: bw=6100MiB/s (6396MB/s), 6100MiB/s-6100MiB/s 
(6396MB/s-6396MB/s), io=16.0GiB (17.2GB), run=2686-2686msec

---- mkfs.btrfs /dev/vdb ..... :0 ----
---- mount /dev/vdb /btrfs ..... :0 ----
---- fio --group_reporting=1 --directory /btrfs --name dioread 
--direct=1 --size=1g --rw=read --runtime=60 --iodepth=1024 --nrfiles=16 
--numjobs=16 | egrep "fio|READ" ..... :0 ----
fio-3.19
    READ: bw=6157MiB/s (6456MB/s), 6157MiB/s-6157MiB/s 
(6456MB/s-6456MB/s), io=16.0GiB (17.2GB), run=2661-2661msec




  reply	other threads:[~2023-01-25  7:43 UTC|newest]

Thread overview: 80+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-01-21  6:49 consolidate btrfs checksumming, repair and bio splitting v4 Christoph Hellwig
2023-01-21  6:49 ` [PATCH 01/34] block: export bio_split_rw Christoph Hellwig
2023-01-21  6:49 ` [PATCH 02/34] btrfs: better document struct btrfs_bio Christoph Hellwig
2023-01-22 10:19   ` Anand Jain
2023-01-23 15:25   ` Johannes Thumshirn
2023-01-21  6:50 ` [PATCH 03/34] btrfs: add a btrfs_inode pointer to " Christoph Hellwig
2023-01-22 10:20   ` Anand Jain
2023-01-23 15:26   ` Johannes Thumshirn
2023-01-23 15:47   ` Johannes Thumshirn
2023-03-07  1:44   ` Qu Wenruo
2023-03-07 14:41     ` Christoph Hellwig
2023-03-07 22:21       ` Qu Wenruo
2023-03-08  6:04         ` Qu Wenruo
2023-03-08 14:28           ` Christoph Hellwig
2023-03-09  0:08             ` Qu Wenruo
2023-03-09  9:31               ` Christoph Hellwig
2023-03-09 10:32                 ` Qu Wenruo
2023-03-09 15:18                   ` Christoph Hellwig
2023-01-21  6:50 ` [PATCH 04/34] btrfs: remove the direct I/O read checksum lookup optimization Christoph Hellwig
2023-01-23 15:59   ` Johannes Thumshirn
2023-01-24 14:55   ` Anand Jain
2023-01-24 19:55     ` Christoph Hellwig
2023-01-25  7:42       ` Anand Jain [this message]
2023-01-21  6:50 ` [PATCH 05/34] btrfs: simplify btrfs_lookup_bio_sums Christoph Hellwig
2023-01-23 16:06   ` Johannes Thumshirn
2023-01-24 15:16   ` Anand Jain
2023-01-21  6:50 ` [PATCH 06/34] btrfs: slightly refactor btrfs_submit_bio Christoph Hellwig
2023-01-23 16:13   ` Johannes Thumshirn
2023-01-24 15:27   ` Anand Jain
2023-01-21  6:50 ` [PATCH 07/34] btrfs: save the bio iter for checksum validation in common code Christoph Hellwig
2023-01-23 16:18   ` Johannes Thumshirn
2023-01-21  6:50 ` [PATCH 08/34] btrfs: pre-load data checksum for reads in btrfs_submit_bio Christoph Hellwig
2023-01-23 16:32   ` Johannes Thumshirn
2023-01-21  6:50 ` [PATCH 09/34] btrfs: add a btrfs_data_csum_ok helper Christoph Hellwig
2023-01-23 16:36   ` Johannes Thumshirn
2023-01-21  6:50 ` [PATCH 10/34] btrfs: handle checksum validation and repair at the storage layer Christoph Hellwig
2023-01-23 16:39   ` Johannes Thumshirn
2023-01-24  6:47     ` Christoph Hellwig
2023-01-24  8:16       ` Johannes Thumshirn
2023-01-21  6:50 ` [PATCH 11/34] btrfs: remove btrfs_bio_free_csum Christoph Hellwig
2023-01-23 16:41   ` Johannes Thumshirn
2023-01-21  6:50 ` [PATCH 12/34] btrfs: remove btrfs_bio_for_each_sector Christoph Hellwig
2023-01-23 16:42   ` Johannes Thumshirn
2023-01-21  6:50 ` [PATCH 13/34] btrfs: remove now unused checksumming helpers Christoph Hellwig
2023-01-23 16:49   ` Johannes Thumshirn
2023-01-23 16:53     ` Christoph Hellwig
2023-01-24  8:33       ` Johannes Thumshirn
2023-01-21  6:50 ` [PATCH 14/34] btrfs: remove the device field in struct btrfs_bio Christoph Hellwig
2023-01-24 11:01   ` Johannes Thumshirn
2023-01-21  6:50 ` [PATCH 15/34] btrfs: remove the io_failure_record infrastructure Christoph Hellwig
2023-01-24 11:04   ` Johannes Thumshirn
2023-01-21  6:50 ` [PATCH 16/34] btrfs: rename the iter field in struct btrfs_bio Christoph Hellwig
2023-01-24 11:09   ` Johannes Thumshirn
2023-01-21  6:50 ` [PATCH 17/34] btrfs: remove the is_metadata flag " Christoph Hellwig
2023-01-24 11:13   ` Johannes Thumshirn
2023-01-21  6:50 ` [PATCH 18/34] btrfs: remove the submit_bio_start helpers Christoph Hellwig
2023-01-24 11:49   ` Johannes Thumshirn
2023-01-21  6:50 ` [PATCH 19/34] btrfs: simplify the btrfs_csum_one_bio calling convention Christoph Hellwig
2023-01-21  6:50 ` [PATCH 20/34] btrfs: handle checksum generation in the storage layer Christoph Hellwig
2023-01-21  6:50 ` [PATCH 21/34] btrfs: handle recording of zoned writes " Christoph Hellwig
2023-01-21  6:50 ` [PATCH 22/34] btrfs: support cloned bios in btree_csum_one_bio Christoph Hellwig
2023-01-21  6:50 ` [PATCH 23/34] btrfs: allow btrfs_submit_bio to split bios Christoph Hellwig
2023-01-25 21:51   ` Josef Bacik
2023-01-26  5:21     ` Christoph Hellwig
2023-01-26 17:43       ` Josef Bacik
2023-01-26 17:46         ` Christoph Hellwig
2023-01-26 18:33           ` David Sterba
2023-01-27  7:02             ` test not in the auto group, was: " Christoph Hellwig
2023-01-21  6:50 ` [PATCH 24/34] btrfs: pass the iomap bio to btrfs_submit_bio Christoph Hellwig
2023-01-21  6:50 ` [PATCH 25/34] btrfs: remove stripe boundary calculation for buffered I/O Christoph Hellwig
2023-01-21  6:50 ` [PATCH 26/34] btrfs: remove stripe boundary calculation for compressed I/O Christoph Hellwig
2023-01-21  6:50 ` [PATCH 27/34] btrfs: remove stripe boundary calculation for encoded I/O Christoph Hellwig
2023-01-21  6:50 ` [PATCH 28/34] btrfs: remove struct btrfs_io_geometry Christoph Hellwig
2023-01-21  6:50 ` [PATCH 29/34] btrfs: remove submit_encoded_read_bio Christoph Hellwig
2023-01-21  6:50 ` [PATCH 30/34] btrfs: remove the fs_info argument to btrfs_submit_bio Christoph Hellwig
2023-01-21  6:50 ` [PATCH 31/34] btrfs: remove now spurious bio submission helpers Christoph Hellwig
2023-01-21  6:50 ` [PATCH 32/34] btrfs: calculate file system wide queue limit for zoned mode Christoph Hellwig
2023-01-21  6:50 ` [PATCH 33/34] btrfs: split zone append bios in btrfs_submit_bio Christoph Hellwig
2023-01-21  6:50 ` [PATCH 34/34] iomap: remove IOMAP_F_ZONE_APPEND Christoph Hellwig
2023-01-24 13:22 ` consolidate btrfs checksumming, repair and bio splitting v4 Christoph Hellwig

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=8d3e27aa-7b65-e08d-7dc1-a8db2d3f1389@oracle.com \
    --to=anand.jain@oracle.com \
    --cc=axboe@kernel.dk \
    --cc=clm@fb.com \
    --cc=damien.lemoal@wdc.com \
    --cc=djwong@kernel.org \
    --cc=dsterba@suse.com \
    --cc=hch@lst.de \
    --cc=johannes.thumshirn@wdc.com \
    --cc=josef@toxicpanda.com \
    --cc=linux-block@vger.kernel.org \
    --cc=linux-btrfs@vger.kernel.org \
    --cc=linux-fsdevel@vger.kernel.org \
    --cc=naohiro.aota@wdc.com \
    --cc=wqu@suse.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