From owner-freebsd-stable@FreeBSD.ORG Fri Apr 4 18:40:32 2014 Return-Path: Delivered-To: freebsd-stable@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:1900:2254:206a::19:1]) (using TLSv1 with cipher ADH-AES256-SHA (256/256 bits)) (No client certificate requested) by hub.freebsd.org (Postfix) with ESMTPS id F2B61764 for ; Fri, 4 Apr 2014 18:40:32 +0000 (UTC) Received: from mail.intertainservices.com (mail.intertainservices.com [69.77.177.114]) by mx1.freebsd.org (Postfix) with ESMTP id CAAA21E2 for ; Fri, 4 Apr 2014 18:40:32 +0000 (UTC) Received: from freebsd.local (unknown [172.16.10.114]) by mail.intertainservices.com (Postfix) with ESMTPSA id 9650356481; Fri, 4 Apr 2014 14:40:20 -0400 (EDT) Message-ID: <533EFC94.7070208@intertainservices.com> Date: Fri, 04 Apr 2014 14:40:20 -0400 From: Mike Jakubik User-Agent: Mozilla/5.0 (X11; FreeBSD amd64; rv:24.0) Gecko/20100101 Thunderbird/24.4.0 MIME-Version: 1.0 To: Steven Hartland , Johan Broman , stable-list freebsd Subject: Re: What's up with the swapping since 10/stable References: In-Reply-To: Content-Type: text/plain; charset=windows-1252; format=flowed Content-Transfer-Encoding: 8bit X-intertainservices-MailScanner-Information: Please contact the ISP for more information X-intertainservices-MailScanner-ID: 9650356481.AD654 X-intertainservices-MailScanner: Found to be clean X-intertainservices-MailScanner-From: mike.jakubik@intertainservices.com X-Spam-Status: No Cc: Matthias Gamsjager X-BeenThere: freebsd-stable@freebsd.org X-Mailman-Version: 2.1.17 Precedence: list List-Id: Production branch of FreeBSD source code List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Fri, 04 Apr 2014 18:40:33 -0000 Hello, I just wanted to mention that i have also observed this behavior. Since i updated from 9 to 10 my swap usage has increased dramatically, but workload has remained the same. This is my desktop pc at work, about a day or two after a reboot swap will be around 50% in use. FreeBSD 10.0-STABLE #0 r263513 Mem: 789M Active, 3842M Inact, 1003M Wired, 115M Cache, 820M Buf, 2156M Free Swap: 1024M Total, 450M Used, 574M Free, 43% Inuse Thanks. On 04/04/14 04:09, Steven Hartland wrote: > We've noticed that swapping is much more agressive in 10.0 than > in 8.3 and have indeed seen it swap processes out when there is > 10GB+ free memory. > > Regards > Steve > ----- Original Message ----- From: "Johan Broman" > To: "stable-list freebsd" > Cc: "Matthias Gamsjager" > Sent: Thursday, April 03, 2014 8:32 PM > Subject: Re: What's up with the swapping since 10/stable > > > Hi! > > I’m seeing the same thing since upgrading to 10/stable. Things seems > to need swap although there is still available memory. I tend not to > use swap on my virtual instances but I’ve seen error messages like > this since upgrading to 10/stable: > > pid 3028 (mysqld), uid 88, was killed: out of swap space > > Mem: 24M Active, 8012K Inact, 109M Wired, 2176K Cache, 69M Buf, 433M Free > > > Looks like there should be enough memory to start mysql… (the above > instance is a t1.micro FreeBSD AMI running on AWS EC2, created by > Colin Percival) > > Something seems to have changed since FreeBSD 9 in terms of memory > manager / page eviction. > > Anyone else seeing this? Is it now impossible to run FreeBSD without a > swap partition (and or file)? This happens on my server as well which > has 8GB RAM and plenty of free RAM… > > I don’t want to start guessing, but perhaps this happens when there is > some memory fragmentation…? I need to verify if this is the case though. > > Thanks > Johan >