From owner-freebsd-hackers Mon Feb 10 11: 2:41 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 5EB7637B405 for ; Mon, 10 Feb 2003 11:02:40 -0800 (PST) Received: from cirb503493.alcatel.com.au (c18609.belrs1.nsw.optusnet.com.au [210.49.80.204]) by mx1.FreeBSD.org (Postfix) with ESMTP id EFF6C43FE3 for ; Mon, 10 Feb 2003 11:02:30 -0800 (PST) (envelope-from peterjeremy@optushome.com.au) Received: from cirb503493.alcatel.com.au (localhost.alcatel.com.au [127.0.0.1]) by cirb503493.alcatel.com.au (8.12.5/8.12.5) with ESMTP id h1AJ2SLZ051575; Tue, 11 Feb 2003 06:02:29 +1100 (EST) (envelope-from jeremyp@cirb503493.alcatel.com.au) Received: (from jeremyp@localhost) by cirb503493.alcatel.com.au (8.12.6/8.12.5/Submit) id h1AJ2Rel051574; Tue, 11 Feb 2003 06:02:27 +1100 (EST) Date: Tue, 11 Feb 2003 06:02:27 +1100 From: Peter Jeremy To: Bogdan TARU Cc: freebsd-hackers@FreeBSD.ORG Subject: Re: gettings snapshots of load spikes Message-ID: <20030210190227.GA50038@cirb503493.alcatel.com.au> References: <20030210182004.G45765-100000@fw.office.icom> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20030210182004.G45765-100000@fw.office.icom> User-Agent: Mutt/1.4i 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 Mon, Feb 10, 2003 at 06:35:43PM +0100, Bogdan TARU wrote: > I am having a real weird problem with a newly installed Dell PowerEdge >2650 which acts as a web (Apache) and mail server(Procmail). The load just >'spikes' sometimes (to 40.00 or so), but immediately starts to go down. ... > One question would be: any idea of how to get a snapshot of the system in >the exact moment when the load sky-rockets? Hack the scheduler so that it drops into DDB if the run queue exceeds say 30. You can then play around inside DDB or force a panic. This probably isn't suitable for a production system. Are the load spikes regular and therefore possibly triggered by cron? Is there anything in either apache or procmail logs that correlates with the spikes? You could try enabling accounting (see acct(5) and accton(8)) which will give you a record of what processes were started. This may give you a clue as to where to start looking. Peter To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message