From owner-freebsd-stable Sun Aug 20 17:45:30 2000 Delivered-To: freebsd-stable@freebsd.org Received: from mail2.uniserve.com (mail2.uniserve.com [204.244.156.10]) by hub.freebsd.org (Postfix) with ESMTP id 3B3DE37B424 for ; Sun, 20 Aug 2000 17:45:28 -0700 (PDT) Received: from shell.uniserve.ca ([204.244.186.218]) by mail2.uniserve.com with esmtp (Exim 3.13 #1) id 13QfiC-0003db-00; Sun, 20 Aug 2000 17:45:12 -0700 Date: Sun, 20 Aug 2000 17:45:07 -0700 (PDT) From: Tom X-Sender: tom@shell.uniserve.ca To: Shawn Barnhart Cc: freebsd-stable@FreeBSD.ORG Subject: Re: SMP and system loads In-Reply-To: <051e01c00afd$4d0e99a0$0102a8c0@k6> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-stable@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG On Sun, 20 Aug 2000, Shawn Barnhart wrote: ... > The previous system had been averaging about 45% CPU utilization, most > of it coming from mrtg monitoring a longish list of devices. After > running most of the day on the SMP-enabled kernel, the load average > doesn't seem to have changed any. Systems load average is not equal to or comparable to CPU utilization. Apples and oranges. > Now I realize that mrtg is mostly a perl script and isn't multithreaded, > but given the fact that I'm running two CPUs, shouldn't the overall load > average be no more than half of what it was? I'm doing same units of > work but have twice as much CPU available for doing them. No. The load average should identical with the same load. The system load average is the average number of processes in the running (or in short term wait). The best thing for your application is to split your mrtg.cfg file into two, and run them through separate instances of mrtg. Tom Uniserve To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message