From owner-freebsd-fs@freebsd.org Tue Jul 14 16:13:06 2015 Return-Path: Delivered-To: freebsd-fs@mailman.ysv.freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:1900:2254:206a::19:1]) by mailman.ysv.freebsd.org (Postfix) with ESMTP id BA8389A18C7 for ; Tue, 14 Jul 2015 16:13:06 +0000 (UTC) (envelope-from csforgeron@gmail.com) Received: from mail-qk0-x234.google.com (mail-qk0-x234.google.com [IPv6:2607:f8b0:400d:c09::234]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (Client CN "smtp.gmail.com", Issuer "Google Internet Authority G2" (verified OK)) by mx1.freebsd.org (Postfix) with ESMTPS id 761A112 for ; Tue, 14 Jul 2015 16:13:06 +0000 (UTC) (envelope-from csforgeron@gmail.com) Received: by qkcl188 with SMTP id l188so9666168qkc.1 for ; Tue, 14 Jul 2015 09:13:05 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20120113; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :cc:content-type; bh=QUf3A/nUs2e39j8G0az62/db2/ajrG9Kb0TwU5mdn8o=; b=qwdek+6MVEgMs4qpH5IVX0le0oNkv3YDJ27f3cmMNOR50n+vycP4Ix893vcT+3QESI xrK4MUymg/Gj5Yddsv+tKajYRyoKRiilMHfLUqJ3GrU1t5NE3yoslv5hJtPihPgGc6U0 +e0OoNencFTkjrb3/p6i6Qui9z+mr4qu1in3dPPV1Lfa8yZs1nhVyQGWgC0jonGjvcs8 lFCyUbrtUBSQijm1uQDJoaprV5vjDxfAwSTQyxh6WdK/ubmuBRgBvY6R/lqrWwZMfYxq QFtHTxYOKeifTxiwGGRW1aGiGd91GQQvTiqvlXr4klVLdi+ZDDd3lmyysOyrzdApmP2M GOow== MIME-Version: 1.0 X-Received: by 10.140.91.66 with SMTP id y60mr62961884qgd.90.1436890385528; Tue, 14 Jul 2015 09:13:05 -0700 (PDT) Received: by 10.96.174.39 with HTTP; Tue, 14 Jul 2015 09:13:05 -0700 (PDT) In-Reply-To: References: <55A3A800.5060904@denninger.net> <55A4D5B7.2030603@freebsd.org> <55A4E5AB.8060909@netlabs.org> Date: Tue, 14 Jul 2015 13:13:05 -0300 Message-ID: Subject: Re: FreeBSD 10.1 Memory Exhaustion From: Christopher Forgeron To: Sean Chittenden Cc: Adrian Gschwend , FreeBSD Filesystems Content-Type: text/plain; charset=UTF-8 X-Content-Filtered-By: Mailman/MimeDel 2.1.20 X-BeenThere: freebsd-fs@freebsd.org X-Mailman-Version: 2.1.20 Precedence: list List-Id: Filesystems List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 14 Jul 2015 16:13:06 -0000 I now have the patch on two of my production machines after early morning crashes that opened up a 'maintenance window'. With crashing happening quite regularly for months now, I think we have progress if I can make it to the end of the week crash-free. On Tue, Jul 14, 2015 at 12:10 PM, Sean Chittenden wrote: > > > ZFS ARC *should* not require those settings, but does currently for mixed > workloads (i.e. databases) in order to be "stable". This illustrates one of my main concerns: - I shouldn't have to tweak/patch FreeBSD to keep it from being unstable. Tweaking to improve performance is perfectly fine with me, that's part of what we do as sysadmins to make the software fit the layout of the required needs. Right now stock FreeBSD is unstable for me, and I don't think my workload is unusual in any way. My smallest production machine is 48 GiB RAM on a 8 TiB Pool, and it still faults like my larger 96GiB RAM machines with larger pools. You can buy a 5 TiB drive for ~$200 - $300 now. It's well within the budget of a home user to have a 10TiB ZFS pool, and if that home user is torrenting video, they are setting up the same situations for crashing. ZFS needs to know when it's consuming swap instead of raw memory, and from what I understand, this patch allows that knowledge, and thus this patch is very important. I'll take a small, possibly performance regression to achieve stability. If we can't come together to help people who's jobs and income literally depend on keeping these boxes running, then can we do it for the common punter trying to torrent movies? :-)