Date: Sun, 1 Jan 2006 23:34:41 -0500 From: Martin Cracauer <cracauer@cons.org> To: Gary Kline <kline@tao.thought.org> Cc: Michael Vince <mv@roq.com>, Martin Cracauer <cracauer@cons.org>, FreeBSD Mailing List <freebsd-questions@freebsd.org> Subject: Re: STressing a new server... Message-ID: <20060101233441.A98774@cons.org> In-Reply-To: <20060102043143.GA88884@thought.org>; from kline@tao.thought.org on Sun, Jan 01, 2006 at 08:31:43PM -0800 References: <20060101212007.GA87257@thought.org> <43B891A3.7040602@roq.com> <20060101232039.A98514@cons.org> <20060102043143.GA88884@thought.org>
next in thread | previous in thread | raw e-mail | index | archive | help
Gary Kline wrote on Sun, Jan 01, 2006 at 08:31:43PM -0800: > On Sun, Jan 01, 2006 at 11:20:40PM -0500, Martin Cracauer wrote: > > For hardware testing, the best is ports/math/mprime > > > > In combination with memtest86, because mprime doesn't sweep all RAM. > > > > If you have several processors, be sure to run several instances of > > mprime (requires copying the whole mprime directory). > > > > Martin > > -- > > Ah, thanks for the tip on mprime. Would the odds of touch more > RAM improve if I ran several nstatiations of mprime at once, > perhaps each differently nice'd? No, prime is best used with nothing else interrupting it, not even switches to other instances of itself. One mprime per CPU. It is a very tightly written assembly program which "cooks" the CPU pretty nicely in torture mode (start with -t). Remember this is hardware test only, it does nothing about OS hickups. Martin -- %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Martin Cracauer <cracauer@cons.org> http://www.cons.org/cracauer/ FreeBSD - where you want to go, today. http://www.freebsd.org/
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20060101233441.A98774>