[NCLUG] ext3 versus ReiserFS?

[Mike Loseke]

Thus spake J. Paul Reed:
> On Wed, 24 Oct 2001, Eric Brunson wrote:
> 
> > Following up my own post:
> >
> > http://www.reiserfs.org/benchmarks/mongo/mongo_res_2.2.18-3.5.30.html
> >
> > If I'm reading the results correctly, reiser beats the crap out of
> > ext3 in most categories of performance.  But that's *if* I'm reading
> > the results right (it's not really clear).
> 
> Speaking of the devil... (look at the posting date):
> 
> http://slashdot.org/article.pl?sid=01/10/23/1621225

 Misleading test results as posted early in the thread, at least when
compared to those goofily organized ones posted here earlier, but
interesting discussion follows to be sure.

 I always find it interesting when people evangelize different JFS setups,
their number one remark is fast fsck/mount times. This is nice or course
but has always seemed misplaced to me. If you're running a big system
where you're dependant on the need for space and oodles of files then,
most likely, it's what you might term as "mission critical". If you're
touting fast mount times when talking about mission critical systems, one
of two things is happenning: 1) you're running some sort of system that
mounts/umounts filesystems left and right locally (where this would be
good), or 2) you have some other problem which is causing yout to reboot
or umount/mount alot. In the case of number 2, maybe you should fix the
problem instead of working around it.

 For my purposes, I'd be more interested in seeing these JFS comparisons
when reading and writing a mix of small files (~50KB) and large files
(~500MB) with the occassional 2GB file thrown in. I'm not running a news
server so I don't give a hoot about a bjillion 1-10KB files.

-- 
   Mike Loseke    | ... Logically incoherent, semantically
 mike at verinet.com | incomprehensible, and legally ... impeccable!