On Mon, 09 Jul 2001, Constantin Loizides wrote:
> I would like to inform you about the status of my project.
> Therefore I have put a page on the web, which gives some details:
> http://www.informatik.uni-frankfurt.de/~loizides/reiserfs/
Interesting. Is there any explanation why Reiserfs performs best with
-no-tails and worst without this option compared to the other filesystems ?
For me packing tails sounds like an excellent way to improve performance,
especially for lots of small files. That it performs much worse than
disabling this code in practice is difficult to understand for me and sounds
more like an implementation bug.
is it difficult/slow to implement tails with the given Linux VFS layer ?
Dirk
