[Xapian-discuss] Filesystems

[Arjen van der Meijden]

On 13-7-2009 17:59 Richard Boulton wrote:
> 2009/7/13 Arjen van der Meijden <acmmailing at tweakers.net>:
>> Executing about 2500 of our slowest queries with a I/O-dependent forced
>> database:
>> ext4    766
>> xfs     768
>> ext3    785
>> ext2    795
>> reiser3 801
>> nilfs   837
>> btrfs   849
> 
> I'm surprised by the range here.  In particular, from what I know of
> their architecture, I wouldn't have expected ext4 and ext3 to be so
> different.

There are a few optimizations in ext4 over ext3 that should also affect 
its read-performance. For instance the extent-support or the improved 
block allocator may matter a bit if blocks are placed more efficiently 
within the underlying stripes.
As these ssd's didn't seem to suffer from performance degradation (and 
ext4 was tested after ext3), we're pretty confident that the results 
represent improvements in the file system.

Best regards,

Arjen