Neil Bothwick writes:

> On Wed, 3 Mar 2010 12:52:55 +0100, Alex Schuster wrote:
> > > The data I've seen indicates that ext2 is fastest, that's what I
> > > use.
> > 
> > I thought the small files of the portage tree especially profit from
> > the notail option in reiserfs?
> 
> They benefit compared with using reiser with tail-packing.

Oh my. I have it the other way around, and never even thought much about 
what this does.

> > Did you change the block size?
> 
> I had to change both the block size and blocks per inode, otherwise I
> would run out of inodes on a 1GB filesystem. You have to admire the
> user-friendliness of ext!

I only wished I could add more inodes after all are out, because this 
happens quite frequently to me. But yes, it's nice I can specify this at 
all.


> > > There's no need for journalling on the portage tree, it's small
> > > enough to fsck quickly and if it does get broken, reformat and
> > > resync.
> > 
> > Would the journaling overhead be noticeable?
> > I also had used ext2 for my portage tree first, then I read somewhere
> > that reiserfs would be the best. BTW, I have distfiles and pkgdir
> > somewhere else, if not the fsck would not be so fast.
> 
> It's certainly noticeable compared with ext3. Many benchmarks do show
> ext2 to be the fastest filesystem, probably because of the lack of
> journalling overhead.

When I saw some, it was maybe 15% difference, and that probably due to 
writes I assume. The portage tree is written during sync only, and then I 
do not care about speed. But would accessing lots and lots of small files 
be slowed down by journaling?

> Like you, I have $DISTDIR and $PKGDIR elsewhere, those files really
> should not be mixed in with the portage tree.
> 
> > Just for fun, I just copied my $PORTDIR into my tmpfs, emerge -DpN
> > @system @world takes between 81 and 53 seconds. With reiserfs, I get
> > 130 seconds first ($PORTDIR was unmounted first and mounted again to
> > clear the caches), and 57 seconds in the second attempt.
> > 
> > I had expected that tmpfs would be even faster. I think I just keep
> > it the way it is now.
> 
> The exact same thought occurred to me. With a local tree to sync from,
> tmpfs seemed a good choice (you could sync it from /etc/conf.d/local)
> but it seems like it is not worth bothering with.

I would need more memory for that, I'm not at amd64 yet. But I probably 
should migrate anyway, and get another 4GB of memory.

> I'll try a reiser3
> filesystem without tail packing to see if it beats ext2.

I backed up my portage tree, re-created the reiserfs partition, and 
mounted without notail option. The same emerge command now takes about 
three minutes... no, on 2nd try it's five. Hmm... ah, clementine is 
indexing files. Why does it do this, I did not change files. Oh, and it 
has indexed all of my /data/mp3, while I only gave it four subfolders to 
index. Why does no audio player just accept my choices for what the 
collection is, and add other stuff?

The next test gives 93 seconds, that's nice.

        Wonko

Reply via email to