Home Messages Index
[Date Prev][Date Next][Thread Prev][Thread Next]
Author IndexDate IndexThread Index

Re: Ext4 Flexes Its Muscles for Early Benchmark

  • Subject: Re: Ext4 Flexes Its Muscles for Early Benchmark
  • From: hackr_d@xxxxxxxxx
  • Date: 22 Oct 2006 09:51:17 -0700
  • Complaints-to: groups-abuse@google.com
  • In-reply-to: <87y7r8mjv9.fsf@geemail.com>
  • Injection-info: b28g2000cwb.googlegroups.com; posting-host=; posting-account=yMVGxgwAAADcjiQJl9DfI98Tt5-VR7wt
  • Newsgroups: comp.os.linux.advocacy
  • Organization: http://groups.google.com
  • References: <1161473647.682267.246670@m7g2000cwm.googlegroups.com> <1161529408_9663@sp6iad.superfeed.net> <87y7r8mjv9.fsf@geemail.com>
  • User-agent: G2/1.0
  • Xref: news.mcc.ac.uk comp.os.linux.advocacy:1169499
Hadron Quark wrote:
> Donn Miller <hackr_d@xxxxxxxxx> writes:
> > Roy Schestowitz wrote:
> >> First benchmarks of the ext4 file system
> >
> > Good job Roy, keep us updated on the news regarding ext3/4 and the
> > rest of the Linux file systems.  I'm interested in finding out more
> Or alternatively just set your home page to any linux news
> web. Faster. More accurate. less discriminating. Real.
> > news.  I was intrigued that the ext3 file system was different
> > internally than ext2, and not simply ext2 with a journal slapped on
> > top of it.  There would have to be some mods made to ext3 in order to
> No you weren't. What do you think ext 3 is?

It's essentially ext2 with a journal.  They're the same internally, but
ext3 uses buffer heads for its directory handling code, while ext2
doesn't.  This is because of the journal.  I'll just have to read the
source for ext2 and ext3 and educate myself, won't I?

> > compensate for the fact that there is a journal, which would otherwise
> > provide some overhead.
> >
> Blah blah blah.

Read above.

I think that overall, ReiserFS is a very good filesystem with many good
ideas.  Ext3 runs faster now in 2.6, because the kernel guys have
managed to tune ext3 to run faster in certain situations.  But Hans
Reiser is a very forward-thinking individual, despite what the trolls
here have done to slander him.  The kernel guys just haven't cooperated
in getting ReiserFS to work as well as it should.  That will come
around, though.   Again, I want to get a hold of some whitepapers on
ReiserFS to find out more.  Yes, I've gone to http://www.namesys.com/,
and i see the "white paper", but it'd be nice to have a PDF format one.

[Date Prev][Date Next][Thread Prev][Thread Next]
Author IndexDate IndexThread Index