New != Faster

[Kris Kennaway]

On Tue, Jun 05, 2007 at 11:06:52PM +0100, Chris wrote:
> On 04/06/07, Kris Kennaway <kris at obsecurity.org> wrote:
> >On Mon, Jun 04, 2007 at 08:54:50PM +0100, Chris wrote:
> >> On 04/06/07, Colin Percival <cperciva at freebsd.org> wrote:
> >> >Tim Daneliuk wrote:
> >> >> Old   2 PIII @600Mhz           768K    26M/sec    4.11-stable/SMP
> >> >> 50-60 min
> >> >> New   Pent D (2 core)@3.2GHz   2G     50M/sec    6.2-stable/SMP
> >> >> 40-50 min
> >> >> Fast  2 Xeon @3GHz             3G    130M/sec    4.11-stable/SMP
> >> >> 8 min
> >> >>
> >> >> Is the difference in speed
> >> >> attributable to 4.11 being faster than 6.2?
> >> >
> >> >Close.  The difference in speed is due to the compiler in 4.11 being
> >> >faster than the compiler in 6.2.  FreeBSD uses the gcc compiler, and
> >> >between FreeBSD 4.11 and FreeBSD 6.2 that has been upgraded from 2.9
> >> >to 3.4.  The general trend each time gcc is upgraded is that it takes
> >> >2x longer to compile code, but produces code which is 5% faster (as a
> >> >result of "working harder" to find optimizations).
> >> >
> >> >FreeBSD 6.2 is faster than FreeBSD 4.11 for almost everything except
> >> >compiling itself. :-)
> >> >
> >> >Colin Percival
> >> >
> >> >_______________________________________________
> >> >freebsd-questions at freebsd.org mailing list
> >> >http://lists.freebsd.org/mailman/listinfo/freebsd-questions
> >> >To unsubscribe, send any mail to
> >> >"freebsd-questions-unsubscribe at freebsd.org"
> >> >
> >>
> >> What about all the following observations?
> >>
> >> slower disk performance especially under QUOTA.
> >
> >s/especially//, unless you have further evidence I don't know about.
> >
> >> both of these have been confirmed numerous times by different people
> >> so sweeping them under the carpet and saying they simply not true
> >> would be wrong.
> >
> >My detailed measurements of disk performance and those of others I am
> >aware of contradicts your claim: 6.x equals or outperforms 4.x on disk
> >I/O (depends on driver) and filesystem I/O.  The only true part of it
> >is the "under QUOTA" part, which as you know from past discussions, is
> >still under Giant in 6.x.  As you also know, there is a patch to
> >address this which is awaiting user testing.  Have you tested it yet?
> >
> >Kris
> >
> Having some hardware coming this week when thats all setup I will have
> a box available for testing patches.

Glad to hear it.  It is kind of irritating that you keep loudly
complaining about how terrible QUOTA performance is but have so far
avoided participating in the solution to that problem.

So, just to confirm, you do not in fact have evidence of poor disk
performance apart from this?

Kris