[HEADS UP] ntfs-3g: better performance with libublio

[Csaba Henk]

On 2007-05-02, Alejandro Pulver <alepulver at FreeBSD.org> wrote:
> I tried to contact fuse4bsd/libublio author, port maintainer, and asked
> in freebsd-hackers@/freebsd-performance@ without response (I am not
> blaming them, perhaps they were busy, I didn't ask correctly or simply
> there was no interest about this there). I don't even know what's the
> OSVERSION value to check this.

Uh, I'm truly sorry, your mail was the victim of my spam filtering
experiments, but now I found that.

> That's why I am asking for testers here. The patch to the current port
> is at people.freebsd.org/~alepulver/fusefs-ntfs.diff, if
> libublio-20070103.tar.gz can't be fetched because the mirrors weren't
> updated get it from: http://people.freebsd.org/~alepulver/).

I see it has been included in the ntfs-3g port since then.

> Without UBLIO_SYNC_IO=3D0 it's slower than without UBLIO, and increasing
> UBLIO_BLOCKSIZE to 1/2/4MB seems a good improvement (also the disk is
> not constantly reading/writing). Increasing UBLIO_ITEMS more than 24
> doesn't seem to increase performance.

I'm glad to hear the good performace of the non-sync case (I won't say
"async", that's a bit more than my non-sync code does).

In fact, I didn't really investigate the non-sync case, because ntfs-3g
developer Szaka said that he would refrain from keeping dirty data in
userspace. But if it's worth to do, probably he will change his
mind. He told me he has plans for looking at the ntfs-3g/libntfs code
with an optimizers' eye, but still there are other things to do.

Csaba