From owner-freebsd-arch@FreeBSD.ORG Mon Jul 4 17:02:40 2005 Return-Path: X-Original-To: arch@freebsd.org Delivered-To: freebsd-arch@FreeBSD.ORG Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 47D1B16A41C; Mon, 4 Jul 2005 17:02:40 +0000 (GMT) (envelope-from dan@dan.emsphone.com) Received: from dan.emsphone.com (dan.emsphone.com [199.67.51.101]) by mx1.FreeBSD.org (Postfix) with ESMTP id 125D443D45; Mon, 4 Jul 2005 17:02:39 +0000 (GMT) (envelope-from dan@dan.emsphone.com) Received: (from dan@localhost) by dan.emsphone.com (8.13.1/8.13.3) id j64H2dvd088731; Mon, 4 Jul 2005 12:02:39 -0500 (CDT) (envelope-from dan) Date: Mon, 4 Jul 2005 12:02:39 -0500 From: Dan Nelson To: Peter Edwards Message-ID: <20050704170239.GQ2392@dan.emsphone.com> References: <20050701132104.GA95135@freefall.freebsd.org> <20050701155757.A36905@fledge.watson.org> <34cb7c8405070311131cd1ca8a@mail.gmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <34cb7c8405070311131cd1ca8a@mail.gmail.com> X-OS: FreeBSD 5.4-STABLE X-message-flag: Outlook Error User-Agent: Mutt/1.5.9i X-Mailman-Approved-At: Tue, 05 Jul 2005 12:35:15 +0000 Cc: Peter Edwards , arch@freebsd.org, Robert Watson Subject: Re: ktrace and KTR_DROP X-BeenThere: freebsd-arch@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Discussion related to FreeBSD architecture List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 04 Jul 2005 17:02:40 -0000 In the last episode (Jul 03), Peter Edwards said: > On 7/1/05, Robert Watson wrote: > > There are two cases where I really run into problems with the > > current model: > > > > (1) When I'm interacting with a slow file system, such as NFS over > > 100mbps, I will always lose records, because it doesn't take > > long for the process to get quite ahead of the write-behind. > > It doesn't even need NFS: syscall throughput is much better than I/O > throughput :-) BTW, I often see dropped records even when tracing to md-backed /tmp. Raising kern.ktrace.request_pool from 100 to 1000 helps but does not eliminate drops. -- Dan Nelson dnelson@allantgroup.com