From owner-freebsd-hackers Fri Mar 14 16:54:45 2003 Delivered-To: freebsd-hackers@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id AEFC737B401 for ; Fri, 14 Mar 2003 16:54:42 -0800 (PST) Received: from office.advantagecom.net (office.advantagecom.net [207.109.186.5]) by mx1.FreeBSD.org (Postfix) with ESMTP id 02BC443F85 for ; Fri, 14 Mar 2003 16:54:42 -0800 (PST) (envelope-from andykinney@advantagecom.net) Received: from scsi-monster (andy.advantagecom.net [207.109.186.200] (may be forged)) by office.advantagecom.net (8.9.3/8.9.3) with ESMTP id QAA26250; Fri, 14 Mar 2003 16:54:35 -0800 From: "Andrew Kinney" Organization: Advantagecom Networks, Inc. To: Julian Elischer Date: Fri, 14 Mar 2003 16:57:18 -0800 MIME-Version: 1.0 Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7BIT Subject: Re: increasing KVA_PAGES and broken pthreads Reply-To: andykinney@advantagecom.net Cc: freebsd-hackers@freebsd.org Message-ID: <3E7209EE.8610.39DAB11@localhost> References: <3E71FB25.27881.363E6F4@localhost> In-reply-to: X-mailer: Pegasus Mail for Win32 (v3.12c) Sender: owner-freebsd-hackers@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG On 14 Mar 2003, at 16:08, Julian Elischer wrote: > > That patch is in the RELENG_4 tree and will be included in 4.8 > Great! Thanks for the info. > > yes it was.. but not in RELENG_4_7 because that is for security > patches. > I'm showing my newbieness here. :-) Apologies. I knew that, but for some reason it didn't occur to me that patches such as this would only occur on the development branches. > > 4.7 is being left behind.. look at 4.8 > I guess I'll just wait for 4.8 to reach "RELEASE" level then and work on reducing the workload of the system in the meantime, though the hardware is nowhere near overloaded. I wish I could track CURRENT, but I'm squeamish about that for a production system such as this. > > IMHO, this issue could be a royal pain in the butt when I start > > working on quad processor systems with 32GB of RAM (not unrealistic > > at this company). > > Well we can't USE 32GB od RAM yet.. I doubt that 4.x will ever be able > to do that (though I could be proven wrong). > Really? I was under the impression that FreeBSD was capable of addressing 8TB of RAM if the hardware supports it. Don't remember which FreeBSD list archive I read that in, but it's not a topic that seems to come up often since most hardware is limited to 4GB of address space. I've got access to hardware that can address 32GB of RAM. Not sure of the exact details of how it works (multiple external memory managers?), but it's a quad Xeon board by SuperMicro. If it's a question of "is there any application that can ever use that much RAM", we're certainly testing the limits here. :-) We're not swapping at all with 4GB, but on several occasions we've gotten close or swapped a few hundred KB. Our two little 2GHz CPUs are humming right along, but most of the time they're better than 60% idle. I imagine that if we pushed the CPUs a bit harder or got hit with a big traffic spike, we'd probably start swapping and want to start thinking about a system that can handle more RAM. Of course, that's assuming that the OS or applications don't break before the hardware gets up to a decent load. Hence, I'm hoping to get a lot of the large memory OS issues resolved (many are by tuning) so we can at least get what we paid for out of the hardware. Sincerely, Andrew Kinney President and Chief Technology Officer Advantagecom Networks, Inc. http://www.advantagecom.net To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message