From: Christoph Hellwig <hch@infradead.org>
To: Dave Chinner <david@fromorbit.com>
Cc: Jan Kara <jack@suse.cz>, linux-scsi <linux-scsi@vger.kernel.org>,
Gluk <git.user@gmail.com>,
Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
xfs@oss.sgi.com, Christoph Hellwig <hch@infradead.org>,
Sergey Meirovich <rathamahata@gmail.com>
Subject: Re: Terrible performance of sequential O_DIRECT 4k writes in SAN environment. ~3 times slower then Solars 10 with the same HBA/Storage.
Date: Mon, 20 Jan 2014 05:58:55 -0800 [thread overview]
Message-ID: <20140120135855.GA26280@infradead.org> (raw)
In-Reply-To: <20140115220721.GY3469@dastard>
On Thu, Jan 16, 2014 at 09:07:21AM +1100, Dave Chinner wrote:
> Yes, I think it can be done relatively simply. We'd have to change
> the code in xfs_file_aio_write_checks() to check whether EOF zeroing
> was required rather than always taking an exclusive lock (for block
> aligned IO at EOF sub-block zeroing isn't required),
That's not even required for supporting aio appends, just a further
optimization for it.
> and then we'd
> have to modify the direct IO code to set the is_async flag
> appropriately. We'd probably need a new flag to say tell the DIO
> code that AIO beyond EOF is OK, but that isn't hard to do....
Yep, need a flag to allow appending writes and then defer them.
> Christoph, are you going to get any time to look at doing this in
> the next few days?
I'll probably need at least another week before I can get to it. If you
wanna pick it up before than feel free.
_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs
WARNING: multiple messages have this Message-ID (diff)
From: Christoph Hellwig <hch@infradead.org>
To: Dave Chinner <david@fromorbit.com>
Cc: Sergey Meirovich <rathamahata@gmail.com>,
Christoph Hellwig <hch@infradead.org>,
xfs@oss.sgi.com, Jan Kara <jack@suse.cz>,
linux-scsi <linux-scsi@vger.kernel.org>,
Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
Gluk <git.user@gmail.com>
Subject: Re: Terrible performance of sequential O_DIRECT 4k writes in SAN environment. ~3 times slower then Solars 10 with the same HBA/Storage.
Date: Mon, 20 Jan 2014 05:58:55 -0800 [thread overview]
Message-ID: <20140120135855.GA26280@infradead.org> (raw)
In-Reply-To: <20140115220721.GY3469@dastard>
On Thu, Jan 16, 2014 at 09:07:21AM +1100, Dave Chinner wrote:
> Yes, I think it can be done relatively simply. We'd have to change
> the code in xfs_file_aio_write_checks() to check whether EOF zeroing
> was required rather than always taking an exclusive lock (for block
> aligned IO at EOF sub-block zeroing isn't required),
That's not even required for supporting aio appends, just a further
optimization for it.
> and then we'd
> have to modify the direct IO code to set the is_async flag
> appropriately. We'd probably need a new flag to say tell the DIO
> code that AIO beyond EOF is OK, but that isn't hard to do....
Yep, need a flag to allow appending writes and then defer them.
> Christoph, are you going to get any time to look at doing this in
> the next few days?
I'll probably need at least another week before I can get to it. If you
wanna pick it up before than feel free.
next prev parent reply other threads:[~2014-01-20 13:58 UTC|newest]
Thread overview: 32+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-01-06 9:38 Terrible performance of sequential O_DIRECT 4k writes in SAN environment. ~3 times slower then Solars 10 with the same HBA/Storage Sergey Meirovich
2014-01-06 20:10 ` Jan Kara
2014-01-07 9:13 ` Sergey Meirovich
2014-01-07 15:58 ` Christoph Hellwig
2014-01-07 18:37 ` Sergey Meirovich
2014-01-08 14:03 ` Christoph Hellwig
2014-01-08 14:43 ` Sergey Meirovich
2014-01-08 15:26 ` Christoph Hellwig
2014-01-08 17:30 ` Sergey Meirovich
2014-01-08 20:55 ` Jan Kara
2014-01-09 10:11 ` Sergey Meirovich
2014-01-10 9:36 ` Jan Kara
2014-01-10 10:36 ` Sergey Meirovich
2014-01-10 10:48 ` Jan Kara
2014-01-10 14:32 ` Sergey Meirovich
2014-01-10 18:14 ` Sergey Meirovich
2014-01-14 13:30 ` Sergey Meirovich
2014-01-14 13:30 ` Sergey Meirovich
2014-01-15 22:07 ` Dave Chinner
2014-01-15 22:07 ` Dave Chinner
2014-01-20 13:58 ` Christoph Hellwig [this message]
2014-01-20 13:58 ` Christoph Hellwig
2014-01-20 22:18 ` Dave Chinner
2014-01-20 22:18 ` Dave Chinner
2014-01-08 1:17 ` Jan Kara
2014-01-08 14:03 ` Christoph Hellwig
2014-01-07 20:57 ` James Smart
2014-01-08 13:57 ` Sergey Meirovich
2014-01-09 19:54 ` Douglas Gilbert
2014-01-09 21:26 ` Sergey Meirovich
2014-01-09 21:43 ` Sergey Meirovich
-- strict thread matches above, loose matches on Subject: below --
2014-01-06 13:16 Sergey Meirovich
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=20140120135855.GA26280@infradead.org \
--to=hch@infradead.org \
--cc=david@fromorbit.com \
--cc=git.user@gmail.com \
--cc=jack@suse.cz \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-scsi@vger.kernel.org \
--cc=rathamahata@gmail.com \
--cc=xfs@oss.sgi.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.