Date: Fri, 15 Mar 2002 01:43:42 -0600 From: "Joe Koberg" <joe@rootnode.com> To: "Dave Uhring" <duhring@charter.net> Cc: <stable@FreeBSD.ORG> Subject: Re: top wrong again? Message-ID: <006901c1cbf5$213ff430$6601000a@flag.rootnode.com> References: <NGBBKILMGLGEDIHMGJANOEDICBAA.sgeine@yahoo.com> <auto-000024329012@dc-mx01.cluster1.charter.net>
next in thread | previous in thread | raw e-mail | index | archive | help
----- Original Message ----- From: "Dave Uhring" <duhring@charter.net> > > With 90% idle, I'd expect to see most of your process running 0%; so > > there doesn't appear to be anything wrong with your top output. > > OK, then try this one. I'm running make buildworld on a 4.5-RELEASE > system with sources CVSup'd about 1/2 hour ago. I've found that many of the processes created during a make buildworld are very short lived, and don't stay around long enough to accumulate processor time. So top(1) shows the ones that are just sitting around long enough for it to capture them during an update. Note the "last pid" number to get an idea of how many processes are created between top's updates. You can always try typing "s 0" to set the update delay to zero in top, but then top is likely to become the most cpu intensive process. Yet, it _is_ fun to watch many cc's and make's being dispatched during a parallel make buildworld. You might want to look at system process accounting with accton(8) and sa(8) to get per-command or per-user totals for cpu usage over time. Have fun, ----------------------------------------- Joe Koberg joe@rootnode.com To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?006901c1cbf5$213ff430$6601000a>