From owner-freebsd-questions@freebsd.org Fri Sep 25 14:01:29 2015 Return-Path: Delivered-To: freebsd-questions@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 8B70FA083CE for ; Fri, 25 Sep 2015 14:01:29 +0000 (UTC) (envelope-from paul@kraus-haus.org) Received: from mail-qk0-f172.google.com (mail-qk0-f172.google.com [209.85.220.172]) (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 494081BBC for ; Fri, 25 Sep 2015 14:01:28 +0000 (UTC) (envelope-from paul@kraus-haus.org) Received: by qkdw123 with SMTP id w123so42312816qkd.0 for ; Fri, 25 Sep 2015 07:01:21 -0700 (PDT) X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:subject:mime-version:content-type:from :in-reply-to:date:cc:content-transfer-encoding:message-id:references :to; bh=wT9OovMEgbRxCvNAFrjNsM1Jwi7Bt7ThlmcznNWxHp8=; b=gvCyDmsqTzFpKRBBavTMUiS5S7FBh4JoEGg0y4SfJnccMU8J17ZkV5GbeUz2dcW4tP HwmJmRvyjc84N7gM+enMxVe97efqsubM1Y2DCa8WUB0JNUI4RlozfY/gJS6a/d3pFxtp Mzn+hamW0HXBNJFHWKd1hIun/+Ay5M5pJHP55BkhfBLNV7kl9i/mME/s26llsEF0s1YW qwN8/4qZpw+w5Ic649ItFBoXBZmE9k089pgkdqA7CI1Ynw6UOzROeKCFaCQc7B7kPby4 +x02W97LJ8FVfqADKVEaTqm8XZsUO9hHHu3c9lNd+pz1TQ/8vwfZHycmH3hnONBMjRam OjeA== X-Gm-Message-State: ALoCoQm/7r0IVX5yNSv8n59xlBJYsETHmZP44ZSG/CnQ6pAP203A/xKPTU/HafYs7Py+oSdordyz X-Received: by 10.55.25.94 with SMTP id k91mr6328334qkh.51.1443189680828; Fri, 25 Sep 2015 07:01:20 -0700 (PDT) Received: from mbp-1.thecreativeadvantage.com (mail.thecreativeadvantage.com. [96.236.20.34]) by smtp.gmail.com with ESMTPSA id r5sm1434931qha.48.2015.09.25.07.01.19 (version=TLSv1 cipher=ECDHE-RSA-RC4-SHA bits=128/128); Fri, 25 Sep 2015 07:01:19 -0700 (PDT) Subject: ZFS ready drives WAS: zfs performance degradation Mime-Version: 1.0 (Mac OS X Mail 7.3 \(1878.6\)) Content-Type: text/plain; charset=windows-1252 From: Paul Kraus In-Reply-To: <5605481D.10902@physics.umn.edu> Date: Fri, 25 Sep 2015 10:01:18 -0400 Cc: freebsd-questions@freebsd.org Content-Transfer-Encoding: quoted-printable Message-Id: <106217D9-F3DB-4DB5-822E-098041B5BC6F@kraus-haus.org> References: <56019211.2050307@dim.lv> <37A37E9D-9D65-4553-BBA2-C5B032163499@kraus-haus.org> <56038054.5060906@dim.lv> <782C9CEF-BE07-4E05-83ED-133B7DA96780@kraus-haus.org> <56040150.90403@dim.lv> <60BF2FC3-0342-46C9-A718-52492303522F@kraus-haus.org> <560412B2.9070905@dim.lv> <8D1FF55C-7068-4AB6-8C0E-B4E64C1BB5FA@kraus-haus.org> <56042209.8040903@dim.lv> <2008181C-F0B5-4581-9D15-11911A1DE41B@kraus-haus.org> <6498A090-A2A2-4580-A148-2BCBF68BF2BF@kraus-haus.org> <5605481D.10902@physics.umn.edu> To: Graham Allan X-Mailer: Apple Mail (2.1878.6) X-BeenThere: freebsd-questions@freebsd.org X-Mailman-Version: 2.1.20 Precedence: list List-Id: User questions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Fri, 25 Sep 2015 14:01:29 -0000 On Sep 25, 2015, at 9:11, Graham Allan wrote: > On 9/24/2015 7:58 PM, Paul Kraus wrote: >> On Sep 24, 2015, at 17:40, CeDeROM wrote: >>=20 >>> For RAID/NAS use ondy WD RED drives family :-) Others WILL break >>> timings with hidden error verify mechanisms.. >>>=20 >>> http://www.wdc.com/red >>=20 >> I assume you are referring to the other =93color=94 drives. I have = had no >> issues with the RE and SE series of Datacenter drives in zpool >> configurations. >=20 > WD Reds are pretty solid, and I have used hundreds of them in ZFS = pools without *apparent* issues, while I would never consider the blue = or even less Green for this. However they're still a low-cost option - = if I'd had the funds I would much rather have used SE or RE! Your comment reminded me=85 one of the big reasons to only use = Enterprise / Datacenter / NAS rated drives for ZFS is the way the = _drive_ handles errors. Many of the consumer drives will retry a failing = READ many, many times. This _can_ lead to timeout issues in the OS and = ZFS. The reasoning here is that for a consumer, getting a good read is = worth the extra time (I have seen reports of up to 30 seconds before = giving up) because the consumer probably does not have any redundancy. = With ZFS (assuming something more than a basic stripe configuration) you = want the drive to return the read error to the OS as fast as it can so = that the OS and ZFS can deal with it. I have also used WD Green and Purple drives with ZFS, but I do not = expect Enterprise grade operation out of them. I also question the economics of the consumer drives, once you take the = 5 year warranty in account. Looking at Newegg for 2 TB 3.5=94 WD drives: Green $79 2-year Purple $85 3-year Red NAS $90 3-year Black $119 5-year SE $130 5-year Red Pro NAS $134 5-year RE $153 5-year So the premium cost for the Red NAS is $11 over the cheapest option. The premium cost for a 5-year warranty (Black, not rated for 24/7 or = NAS, a high end desktop drive) is $40. The Premium for the cheapest Datacenter drive (SE) is $51, or more than = a 50% increase in cost. But the warranty is more than twice as long = (2-year vs. 5-year). In my experience, most 5-year warranty drives fail in some way during = the warranty period. This is especially true of Seagate. On my home = system, 5 out of 6 Seagate ES.2 series drives failed within 5 years, the = last one failed within 6 months of the warranty expiration. Half of my = HGST drives have failed under warranty (so far, they have not all hit = end of warranty yet), none of my WD RE or SE series have failed, but = they are the youngest drives in my collection. So part of what I am paying for with the Datacenter drives is the = knowledge that I will NOT have to pay to replace that drive for 5 years. -- Paul Kraus paul@kraus-haus.org