On Thu, Sep 27, 2001 at 06:42:09PM -0500, Bob Tanner wrote:
> Quoting Jon Erickson (jon.erickson at neicoltech.org):
> > >The testing I have done, shows ext2 faster the Reiserfs, and on a cache fs, I
> > >don't think journaling is necessary or needed (IMHO).
> > 
> > That's interesting to hear.  I've read on a couple web sites 
> > (http://www.swelltech.com/pengies/joe/benchmarks.html   is one of them) 
> > that had benchmarked ext2 and Reiserfs cache partitions and they saw 
> > significant performance increases with Reiserfs.  I'd be interested in 
> > seeing your results if you have them documented.
> 
> I tried this on a fs where I do all my compiles. This is under Java using both
> make and ant, so I don't know if .java and .class files constitute "small"
> files.
> 
> I got the raw data, let me see of I can make something with gnuplot.

What? Your benchmarked filesystems by compiling java sources?

What has java compilation to do with filesystem access? Have you did a vmstat
to see how much time is spent in i/o and how much time actually compiling?

Your "benchmark" is totally flawed.

florin

-- 

"If it's not broken, let's fix it till it is."

41A9 2BDE 8E11 F1C5 87A6  03EE 34B3 E075 3B90 DFE4
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 230 bytes
Desc: not available
Url : http://shadowknight.real-time.com/pipermail/tclug-list/attachments/20010927/ffdeb60b/attachment.pgp