[PLUG] When tmpfs==good and ext3==bad :-)

[Dan Young]

On Wed, 14 May 2003, Paul Heinlein wrote:

> Obviously, a filesystem living in RAM is going to outperform one on 
> disk, but I don't think that explains a 5x to 7x speed difference.

I always thought RAM was an order of magnitude faster than disks. Is that 
a bad assumption?
 
> It seems to me that the journaling overhead becomes unbearable on all
> those little atomic db updates.

Have you thought about mounting the ext3 partitions in data=ordered or 
data=writeback modes? What about another filesystem? I've heard 
anecdotally that ReiserFS is fast for bazillions of small files.

-Dan Young
-Parkrose School District