From: Chris Mason <chris.mason@oracle.com>
To: Boaz Harrosh <bharrosh@panasas.com>
Cc: "Ted Ts'o" <tytso@mit.edu>, Zach Brown <zab@zabbo.net>,
Eric Sandeen <sandeen@redhat.com>,
linux-fsdevel@vger.kernel.org, linux-ext4@vger.kernel.org
Subject: Re: [PATCH, RFC] Don't do page stablization if !CONFIG_BLKDEV_INTEGRITY
Date: Thu, 8 Mar 2012 15:37:41 -0500 [thread overview]
Message-ID: <20120308203741.GE29510@shiny> (raw)
In-Reply-To: <4F59148A.4070001@panasas.com>
On Thu, Mar 08, 2012 at 12:20:26PM -0800, Boaz Harrosh wrote:
> On 03/08/2012 10:09 AM, Chris Mason wrote:
> >
> > But, why are we writeback for a second or more? Aren't there other
> > parts of this we would want to fix as well?
> >
> > I'm not against only turning on stable pages when they are needed, but
> > the code that isn't the default tends to be somewhat less used. So it
> > does increase testing burden when we do want stable pages, and it tends
> > to make for awkward bugs that are hard to reproduce because someone
> > neglects to mention it.
> >
> > IMHO it's much more important to nail down the 2 second writeback
> > latency. That's not good.
> >
>
> I think I understand this one. It's do to the sync nature introduced
> by page_waiting in mkwrite.
Pages go from dirty to writeback for a few reasons. Background
writeout, or O_DIRECT or someone running sync
background writeout shouldn't be queueing up so much work that
synchronous writeout has a 2 second delay.
If the latencies are coming from something that was run through
fsync...well there's not too much we can do about that. The problem is
that our page_mkwrite call isn't starting the IO it is just waiting on
it, so we can't bump the priority on it.
-chris
next prev parent reply other threads:[~2012-03-08 20:37 UTC|newest]
Thread overview: 35+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-03-07 23:40 [PATCH, RFC] Don't do page stablization if !CONFIG_BLKDEV_INTEGRITY Theodore Ts'o
2012-03-07 23:54 ` Eric Sandeen
2012-03-08 0:05 ` Darrick J. Wong
2012-03-08 2:18 ` Darrick J. Wong
2012-03-08 3:00 ` Boaz Harrosh
2012-03-08 3:21 ` Boaz Harrosh
2012-03-08 2:39 ` Zach Brown
2012-03-08 15:54 ` Ted Ts'o
2012-03-08 18:09 ` Chris Mason
2012-03-08 20:20 ` Boaz Harrosh
2012-03-08 20:37 ` Chris Mason [this message]
2012-03-08 20:42 ` Jeff Moyer
2012-03-08 20:55 ` Chris Mason
2012-03-08 21:12 ` Ted Ts'o
2012-03-08 21:20 ` Chris Mason
2012-03-09 8:11 ` Dave Chinner
2012-03-08 20:50 ` Boaz Harrosh
2012-03-08 23:32 ` Dave Chinner
2012-03-08 21:24 ` Ted Ts'o
2012-03-08 21:38 ` Chris Mason
2012-03-08 21:41 ` Ted Ts'o
2012-03-09 1:02 ` Chris Mason
2012-03-09 1:08 ` Martin K. Petersen
2012-03-09 16:20 ` Ted Ts'o
2012-03-08 21:52 ` Boaz Harrosh
2012-03-08 0:23 ` Boaz Harrosh
2012-03-08 3:45 ` Martin K. Petersen
2012-03-08 4:37 ` Boaz Harrosh
2012-03-08 6:27 ` Sage Weil
2012-03-08 15:43 ` Ted Ts'o
2012-03-08 16:36 ` Martin K. Petersen
2012-03-08 16:43 ` Sage Weil
2012-03-15 2:10 ` Andy Lutomirski
2012-03-15 4:46 ` Boaz Harrosh
2012-03-15 5:02 ` Andy Lutomirski
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=20120308203741.GE29510@shiny \
--to=chris.mason@oracle.com \
--cc=bharrosh@panasas.com \
--cc=linux-ext4@vger.kernel.org \
--cc=linux-fsdevel@vger.kernel.org \
--cc=sandeen@redhat.com \
--cc=tytso@mit.edu \
--cc=zab@zabbo.net \
/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).