On Mon, 27 Aug 2001 19:50:25 +0100, Rasputin wrote:
> > > It made perfromance on BSD suck, since every byte written caused an
> > > fsync() call, despite the fact the at the time FreeBSD mounted it's
> > > disks synchronously.
> >
> > Any BSD administrator worth her salt knows about Soft Updates by this
> > stage.
>
> That makes no difference if an app insists on calling fsync()
> unneccessarily, does it? You still take performance hits.
Hmmm, I take back my comments. I thought soft updates helped in this
scenario, but after some consultation with Peter Wemm, I stand corrected.
In fact, Peter seems to think that soft updates actually exacerbates the
situation for "badly behaved" applications that fsync() "too much".
