[ On Tuesday, May 6, 2003 at 22:41:08 (+1000), Paul Ripke wrote: ]
> Subject: Re: some performance diff with getc()/putc() between FreeBSD and NetBSD?
>
> OK, what concerns me is those "random seeks/sec" rates. There's a huge
> difference there, and that could hurt other stats. It shouldn't be CPU
> bound (obviously), the only thing that comes to mind is in disk
> I/O scheduling - disksort. I wonder what the difference is.

Yes, that concerns me a bit too.

I'm not sure how to quantify the difference in seeks/sec though.

Bonnie does this part of the test by starting four child processes which
then each do 4000 random lseeks after which they reads a block of data
reads, then for 10% of the operations they also change the data and
write it back.  However I'm not sure they do as many total operations as
they should, even on a very large file, to obliterate the differences in
the buffer cache, especially not on such a large system.

I was going to add options to bonnie to allow me to change the number of
operations and the number of child processes (and perhaps even the write
ratio), but I didn't get a chance to do this while FreeBSD was still on
the system.

Much more testing would have to be done, and also with other benchmarks,
before anything definite could be said about this difference, but I'm
not going to get the chance to reload FreeBSD onto this particular
machine.

-- 
								Greg A. Woods

+1 416 218-0098;            <g.a.woods@ieee.org>;           <woods@robohack.ca>
Planix, Inc. <woods@planix.com>; VE3TCP; Secrets of the Weird <woods@weird.com>