Date: Mon, 26 Jun 2000 22:43:45 -0500 (CDT) From: James Wyatt <jwyatt@rwsystems.net> To: Max Clark <max.clark@emind.com> Cc: freebsd-isp@freebsd.org Subject: Re: interpreting uptime Message-ID: <Pine.BSF.4.10.10006261648160.52036-100000@bsdie.rwsystems.net> In-Reply-To: <004001bfdfab$9c849700$950110ac@emindnfzmj9j9m>
next in thread | previous in thread | raw e-mail | index | archive | help
As I understand it, it is the number of jobs waiting to run on a CPU. When I first used it in 1988, it was the 5/10/15 min averages. The current man page says it's the 1/5/15 minute averages. I've seen a machine dragging to a halt at 18 to 25 before we had to reboot it to stop the process storm. After 18, we couldn't issue careful kills as fast as the rogue processes were creating children, hence reboot and "everyone out of the pool"... On multiprocessor machines, you can mentally divide by the number of CPUs to get a number that feels the same on a single CPU box. I've used one with 4 CPUs and a load of 10 and it was pretty zippy; it just had hundreds of HTTP processes from a large user load. Hope this helps - Jy@ On Mon, 26 Jun 2000, Max Clark wrote: > Okay- really basic question. How do you interpert the values that uptime > gives you? I know that the first value is current, five minutes, and fifteen > minutes. But what do the number values mean? At what point should I become > concerned with the machine? To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-isp" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.BSF.4.10.10006261648160.52036-100000>