Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 19 Apr 2004 21:43:38 -0300 (ADT)
From:      "Marc G. Fournier" <scrappy@hub.org>
To:        Willem Jan Withagen <wjw@withagen.nl>
Cc:        freebsd-current@freebsd.org
Subject:   Re: Is -current really this slow, or do I have something mis-configured?
Message-ID:  <20040419214244.D780@ganymede.hub.org>
In-Reply-To: <020c01c42667$8d70bf80$471b3dd4@dual>
References:  <20040419175924.Y971@ganymede.hub.org> <020c01c42667$8d70bf80$471b3dd4@dual>

next in thread | previous in thread | raw e-mail | index | archive | help

Stupid me, I had commented out KERNCONF in my /etc/make.conf at some
point, so I've been building a GENERIC kernel without realizing it :(
Just re-built a new kernel and installed, which will hopefully fix that up
for tomorrow when I get back to the office *cross fingers*

On Tue, 20 Apr 2004, Willem Jan Withagen wrote:

>
> ----- Original Message -----
> From: "Marc G. Fournier" <scrappy@hub.org>
> To: <freebsd-current@freebsd.org>
> Sent: Monday, April 19, 2004 11:04 PM
> Subject: Is -current really this slow, or do I have something mis-configured?
>
>
> >
> > I'm running -CURRENT on my desktop ... the machine is a P4, no HTT, and
> > 512Meg of RAM ... and I swear the machine is less responsive then my
> > laptop running on a Celeron 333 with ~192 Meg of RAM ... both are IDE
> > drives, both are runing X and both are -current ...
>
> If you are running -CURRENT with WITNESS and all other assert stuff then
> there could be a serious "penalty" you pay. And there is a good reason for it:
>
> Nota Bene:
> Turn off WITNESS and INVARIANTS only
> when benchmarking or for production systems
> Robert Watson writes: "We turn them off in releases,
> and once 5.x becomes 5-stable, we'll turn it off by default also. However,
> they're invaluable tools when debugging the development system, so we have
> them on in the development branch by default. I would encourage people to
> generally run with them turned on unless performance of a system requires
> them to be off, as it really helps the debugging process, as well as
> helping to identify locking problems as the system evolves."
>
>
> You might want to look at the stats I have for a simple bonnie run with and
> without
> the WITNESS stuff. Block Write throughput is not much different, but there is a
> huge
> difference when "fetching" data which was still in memory. I could very well
> imagine
> that a similar effect would impact your feel of responsiveness.
>     http://freebee.digiware.nl/FreeBSD/NFS-performance/#bonnie-local
>
> --WjW
>
>

----
Marc G. Fournier           Hub.Org Networking Services (http://www.hub.org)
Email: scrappy@hub.org           Yahoo!: yscrappy              ICQ: 7615664



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20040419214244.D780>