From: Ming Lin <mlin@kernel.org>
To: Ming Lin <mlin@kernel.org>
Cc: Alasdair G Kergon <agk@redhat.com>,
Mike Snitzer <snitzer@redhat.com>,
lkml <linux-kernel@vger.kernel.org>,
Christoph Hellwig <hch@lst.de>,
Kent Overstreet <kent.overstreet@gmail.com>,
Jens Axboe <axboe@kernel.dk>, Dongsu Park <dpark@posteo.net>,
Christoph Hellwig <hch@infradead.org>,
Al Viro <viro@zeniv.linux.org.uk>,
Ming Lei <ming.lei@canonical.com>, Neil Brown <neilb@suse.de>,
dm-devel@redhat.com, Lars Ellenberg <drbd-dev@lists.linbit.com>,
drbd-user@lists.linbit.com, Jiri Kosina <jkosina@suse.cz>,
Geoff Levand <geoff@infradead.org>, Jim Paris <jim@jtan.com>,
Joshua Morris <josh.h.morris@us.ibm.com>,
Philip Kelleher <pjk1939@linux.vnet.ibm.com>,
Minchan Kim <minchan@kernel.org>, Nitin Gupta <ngupta@vflare.org>,
Oleg Drokin <oleg.drokin@intel.com>,
Andreas Dilger <andreas.dilger@intel.com>
Subject: Re: [PATCH v4 01/11] block: make generic_make_request handle arbitrarily sized bios
Date: Tue, 2 Jun 2015 13:59:06 -0700 [thread overview]
Message-ID: <CAF1ivSY_7M1OhrMuXg0OKu7BPy=SbTHSHCk+2q6vfEVgvJL8YA@mail.gmail.com> (raw)
In-Reply-To: <1433138551.11778.4.camel@hasee>
On Sun, May 31, 2015 at 11:02 PM, Ming Lin <mlin@kernel.org> wrote:
> On Thu, 2015-05-28 at 01:36 +0100, Alasdair G Kergon wrote:
>> On Wed, May 27, 2015 at 04:42:44PM -0700, Ming Lin wrote:
>> > Here are fio results of XFS on a DM stripped target with 2 SSDs + 1 HDD.
>> > Does it make sense?
>>
>> To stripe across devices with different characteristics?
>>
>> Some suggestions.
>>
>> Prepare 3 kernels.
>> O - Old kernel.
>> M - Old kernel with merge_bvec_fn disabled.
>> N - New kernel.
>>
>> You're trying to search for counter-examples to the hypothesis that
>> "Kernel N always outperforms Kernel O". Then if you find any, trying
>> to show either that the performance impediment is small enough that
>> it doesn't matter or that the cases are sufficiently rare or obscure
>> that they may be ignored because of the greater benefits of N in much more
>> common cases.
>>
>> (1) You're looking to set up configurations where kernel O performs noticeably
>> better than M. Then you're comparing the performance of O and N in those
>> situations.
>>
>> (2) You're looking at other sensible configurations where O and M have
>> similar performance, and comparing that with the performance of N.
>
> I didn't find case (1).
>
> But the important thing for this series is to simplify block layer
> based on immutable biovecs. I don't expect performance improvement.
>
> Here is the changes statistics.
>
> "68 files changed, 336 insertions(+), 1331 deletions(-)"
>
> I run below 3 test cases to make sure it didn't bring any regressions.
> Test environment: 2 NVMe drives on 2 sockets server.
> Each case run for 30 minutes.
>
> 2) btrfs radi0
>
> mkfs.btrfs -f -d raid0 /dev/nvme0n1 /dev/nvme1n1
> mount /dev/nvme0n1 /mnt
>
> Then run 8K read.
>
> [global]
> ioengine=libaio
> iodepth=64
> direct=1
> runtime=1800
> time_based
> group_reporting
> numjobs=4
> rw=read
>
> [job1]
> bs=8K
> directory=/mnt
> size=1G
>
> 2) ext4 on MD raid5
>
> mdadm --create /dev/md0 --level=5 --raid-devices=2 /dev/nvme0n1 /dev/nvme1n1
> mkfs.ext4 /dev/md0
> mount /dev/md0 /mnt
>
> fio script same as btrfs test
>
> 3) xfs on DM stripped target
>
> pvcreate /dev/nvme0n1 /dev/nvme1n1
> vgcreate striped_vol_group /dev/nvme0n1 /dev/nvme1n1
> lvcreate -i2 -I4 -L250G -nstriped_logical_volume striped_vol_group
> mkfs.xfs -f /dev/striped_vol_group/striped_logical_volume
> mount /dev/striped_vol_group/striped_logical_volume /mnt
>
> fio script same as btrfs test
>
> ------
>
> Results:
>
> 4.1-rc4 4.1-rc4-patched
> btrfs 1818.6MB/s 1874.1MB/s
> ext4 717307KB/s 714030KB/s
> xfs 1396.6MB/s 1398.6MB/s
Hi Alasdair & Mike,
Would you like these numbers?
I'd like to address your concerns to move forward.
Thanks.
next prev parent reply other threads:[~2015-06-02 20:59 UTC|newest]
Thread overview: 62+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-05-22 18:18 [PATCH v4 00/11] simplify block layer based on immutable biovecs Ming Lin
2015-05-22 18:18 ` [PATCH v4 01/11] block: make generic_make_request handle arbitrarily sized bios Ming Lin
2015-05-25 5:46 ` NeilBrown
2015-05-26 14:36 ` Mike Snitzer
2015-05-26 15:02 ` Ming Lin
2015-05-26 15:34 ` Alasdair G Kergon
2015-05-26 23:06 ` NeilBrown
2015-05-27 0:40 ` [dm-devel] " Alasdair G Kergon
2015-05-27 8:20 ` Christoph Hellwig
2015-05-26 16:04 ` Mike Snitzer
2015-05-26 17:17 ` Ming Lin
2015-05-27 23:42 ` Ming Lin
2015-05-28 0:36 ` Alasdair G Kergon
2015-05-28 5:54 ` Ming Lin
2015-05-29 7:05 ` Ming Lin
2015-05-29 15:15 ` Mike Snitzer
2015-06-01 6:02 ` Ming Lin
2015-06-02 20:59 ` Ming Lin [this message]
2015-06-04 21:06 ` Mike Snitzer
2015-06-04 22:21 ` Ming Lin
2015-06-05 0:06 ` Mike Snitzer
2015-06-05 5:21 ` Ming Lin
2015-06-09 6:09 ` Ming Lin
2015-06-10 21:20 ` Ming Lin
2015-06-10 21:46 ` Mike Snitzer
2015-06-10 22:06 ` Ming Lin
2015-06-12 5:49 ` Ming Lin
2015-06-18 5:27 ` Ming Lin
2015-05-22 18:18 ` [PATCH v4 02/11] block: simplify bio_add_page() Ming Lin
2015-05-22 18:18 ` [PATCH v4 03/11] bcache: remove driver private bio splitting code Ming Lin
2015-05-22 18:18 ` [PATCH v4 04/11] btrfs: remove bio splitting and merge_bvec_fn() calls Ming Lin
2015-05-22 18:18 ` [PATCH v4 05/11] block: remove split code in blkdev_issue_discard Ming Lin
2015-05-22 18:18 ` [PATCH v4 06/11] md/raid5: get rid of bio_fits_rdev() Ming Lin
2015-05-25 5:48 ` NeilBrown
2015-05-25 7:03 ` Ming Lin
2015-05-25 7:54 ` NeilBrown
2015-05-25 14:17 ` Christoph Hellwig
2015-05-26 14:33 ` Ming Lin
2015-05-26 22:32 ` Ming Lin
2015-05-26 23:03 ` NeilBrown
2015-05-26 23:42 ` Ming Lin
2015-05-27 0:38 ` NeilBrown
2015-05-27 8:15 ` Christoph Hellwig
2015-05-22 18:18 ` [PATCH v4 07/11] md/raid5: split bio for chunk_aligned_read Ming Lin
2015-05-22 18:18 ` [PATCH v4 08/11] block: kill merge_bvec_fn() completely Ming Lin
2015-05-25 5:49 ` NeilBrown
2015-05-25 14:04 ` Christoph Hellwig
2015-05-25 15:02 ` Ilya Dryomov
2015-05-25 15:08 ` Christoph Hellwig
2015-05-25 15:19 ` Ilya Dryomov
2015-05-25 15:35 ` Alex Elder
2015-05-22 18:18 ` [PATCH v4 09/11] fs: use helper bio_add_page() instead of open coding on bi_io_vec Ming Lin
2015-05-22 18:18 ` [PATCH v4 10/11] block: remove bio_get_nr_vecs() Ming Lin
2015-05-22 18:18 ` [PATCH v4 11/11] Documentation: update notes in biovecs about arbitrarily sized bios Ming Lin
2015-05-23 14:15 ` [PATCH v4 00/11] simplify block layer based on immutable biovecs Christoph Hellwig
2015-05-24 7:37 ` Ming Lin
2015-05-25 13:51 ` Christoph Hellwig
2015-05-29 6:39 ` Ming Lin
2015-06-01 6:15 ` Ming Lin
2015-06-03 6:57 ` Christoph Hellwig
2015-06-03 13:28 ` Jeff Moyer
2015-06-03 17:06 ` Ming Lin
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='CAF1ivSY_7M1OhrMuXg0OKu7BPy=SbTHSHCk+2q6vfEVgvJL8YA@mail.gmail.com' \
--to=mlin@kernel.org \
--cc=agk@redhat.com \
--cc=andreas.dilger@intel.com \
--cc=axboe@kernel.dk \
--cc=dm-devel@redhat.com \
--cc=dpark@posteo.net \
--cc=drbd-dev@lists.linbit.com \
--cc=drbd-user@lists.linbit.com \
--cc=geoff@infradead.org \
--cc=hch@infradead.org \
--cc=hch@lst.de \
--cc=jim@jtan.com \
--cc=jkosina@suse.cz \
--cc=josh.h.morris@us.ibm.com \
--cc=kent.overstreet@gmail.com \
--cc=linux-kernel@vger.kernel.org \
--cc=minchan@kernel.org \
--cc=ming.lei@canonical.com \
--cc=neilb@suse.de \
--cc=ngupta@vflare.org \
--cc=oleg.drokin@intel.com \
--cc=pjk1939@linux.vnet.ibm.com \
--cc=snitzer@redhat.com \
--cc=viro@zeniv.linux.org.uk \
/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).