Date: Fri, 7 Apr 2006 10:16:22 +0200 (CEST) From: patrick <patrick@barmentlo.nl> To: Robert Leftwich <freebsd@rtl.fmailbox.com> Cc: freebsd-amd64@freebsd.org Subject: Re: extreme mem usage under amd64 arch ? Message-ID: <20060407101400.R28178@gandalf.cyberwizards.nl> In-Reply-To: <443613D0.80801@rtl.fmailbox.com> References: <20060407075301.I21526@gandalf.cyberwizards.nl> <443613D0.80801@rtl.fmailbox.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Fri, 7 Apr 2006, Robert Leftwich wrote: Thanks for your reply, starting to get real concerned now. we are on the moment of migration and taking this server life.. thinking what is wise.. Even consuring to 'downgrade' to 1386 at this moment. it this do-able on this amd system, a build world , would i be able to do a install world/kernels without crashing on the way??? Regards, Patrick >> Hope someone can shine a light for me on this, > > Sorry, I can't do a lot more than 'me too' > > I reported similar issues on an asus a8n-sli with dual athlon's running > STABLE 6 some time back w/o anyone being able to shed light on it. > > Do you have the ability to run the latest 6.1 beta? If so, and your > experience is like mine, the problem goes away. I'm running the exact same > setup on 6.1b4 as I was on 6 stable and I don't see anywhere near the same > amount of mem usage as I did before - in fact I've been able to triple some > of my Postgres buffer settings with the 4G of ram not even 1/2 fully > utilised. > > I saw very similar memory sizes with httpd as you reported before I switched > to lighttpd (which interestingly did not have the same problem on STABLE 6 - > always using around 17M) but some Python processes I was running regularly > maxed out at over 1.4G, now it's down in the 400M range doing the same > processing on significantly more data. > > HTH > > Robert >
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20060407101400.R28178>