strict allocate

[Björn Jacke]

On 2009-12-01 at 10:11 -0800 Jeremy Allison sent off:
> > and it avoids samba's misbehaviour when quota limits are reached.
> 
> Which is a rare case.

even rare failure cases are bad, aren't they?


> > it fixes ext3 - and with ext4 we'll see no performance penalty at all when
> > posix_fallocate/fallocate would be used.
> 
> Haha ! Oh ye of great faith :-).

:-)


> > see above. With filesystems like ext4 or xfs we have no overhead for
> > preallocation of space.
> 
> Not true. Look at the glibc implementation of posix_fallocate.

and fallocate in glibc 2.10.


> > correct me if I'm wrong but it's mostly a workaround for the extreme
> > framenatations problem. it doesn't fix the we-run-out-of-quota situation and
> > the fallocate for ext4 and xfs is also not yet there.
> 
> It doesn't fix the problem that strict allocate = true does,
> but there's a reason strict allocate = true is not default.
> 
> You haven't convinced me of the need to change it. Numbers
> and benchmarks would do that.

take a look at these numbers:

http://fedorapeople.org/gitweb?p=amitshah/public_git/alloc-perf.git;a=blob_plain;f=results.txt;hb=HEAD

http://log.amitshah.net/2009/03/comparison-of-file-systems-and-speeding.html

Cheers
Björn