xfs
[Top] [All Lists]

Re: ext4 writepages is making tiny bios?

To: Dave Chinner <david@xxxxxxxxxxxxx>
Subject: Re: ext4 writepages is making tiny bios?
From: Christoph Hellwig <hch@xxxxxxxxxxxxx>
Date: Thu, 3 Sep 2009 12:42:09 -0400
Cc: Christoph Hellwig <hch@xxxxxxxxxxxxx>, Theodore Tso <tytso@xxxxxxx>, Chris Mason <chris.mason@xxxxxxxxxx>, linux-ext4@xxxxxxxxxxxxxxx, linux-fsdevel@xxxxxxxxxxxxxxx, xfs@xxxxxxxxxxx
In-reply-to: <20090903055201.GA7146@xxxxxxxxxxxxxxxx>
References: <20090901184450.GB7885@think> <20090901205744.GE6996@xxxxxxx> <20090901212740.GA9930@xxxxxxxxxxxxx> <20090903055201.GA7146@xxxxxxxxxxxxxxxx>
User-agent: Mutt/1.5.19 (2009-01-05)
On Thu, Sep 03, 2009 at 03:52:01PM +1000, Dave Chinner wrote:
> > XFS did the mistake of trusting the VM, while everyone more or less
> > overrode it.  Removing all those checks and writing out much larger
> > data fixes it with a relatively small patch:
> > 
> >     http://verein.lst.de/~hch/xfs/xfs-writeback-scaling
> 
> Careful:
> 
> -     tloff = min(tlast, startpage->index + 64);
> +     tloff = min(tlast, startpage->index + 8192);
> 
> That will cause 64k page machines to try to write back 512MB at a
> time. This will re-introduce similar to the behaviour in sles9 where
> writeback would only terminate at the end of an extent (because the
> mapping end wasn't capped like above).

Pretty good point, any applies to all the different things we discussed
recently.  Ted, should be maybe introduce a max_writeback_mb instead of
the max_writeback_pages in the VM, too?

<Prev in Thread] Current Thread [Next in Thread>