Re: [GIT PULL] ext4 updates for 3.9

[Dave Chinner <david@fromorbit.com>]

On Wed, Feb 27, 2013 at 02:29:07PM -0500, Theodore Ts'o wrote:
> On Wed, Feb 27, 2013 at 02:19:23PM -0500, Dave Jones wrote:
> > 
> > Looks like it's fixed here too.
> > 
> > How did this make it through -next without anyone hitting it ?
> 
> > Is anyone running xfstests or similar on linux-next regularly ?
> 
> I run xfstests on the ext4 tree, and I ran it on ext4 plus Linus's tip
> before I submitted a pull request.  The problem is that XFSTESTS is
> S-L-O-W if you use large partitions, so typically I use a 5GB
> partition sizes for my test runs.

This isn't the case for XFS. I typically see 1TB scratch devices
only being ~10-20% slower than 10GB scratch devices, and 10TB only
being a little slower than 1TB scratch devices. I have to use sparse
devices and --large-fs for 100TB filesystem testing, so I can't
directly compare the speeds to those that I run on physical devices.
However I can say that it isn't significantly slower than using
small scratch devices...

> So what we probably need to do is to have a separate set of tests
> using a loopback mount, and perhaps an artificially created file
> system which has a large percentage of the blocks in the middle of the
> file system busied out, to make efficient testing of these sorts of

That's exactly what the --large-fs patch set I posted months ago does
for ext4 - it uses fallocate() to fill all but 50GB of the large
filesystem without actually writing any data and runs the standard
tests in the remaining unused space.

However, last time I tested ext4 with this patchset (when I posted
the patches months ago), multi-TB preallocation on ext4 was still too
slow to make it practical for testing on devices larger than 2-3TB.
Perhaps it would make testing 1-2TB ext4 filesystems fast enough
that you could do it regularly...

Cheers,

Dave.
-- 
Dave Chinner
david@fromorbit.com