From owner-freebsd-questions@FreeBSD.ORG Sun Apr 10 14:06:21 2011 Return-Path: Delivered-To: freebsd-questions@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 098D2106566C for ; Sun, 10 Apr 2011 14:06:21 +0000 (UTC) (envelope-from DStaal@usa.net) Received: from mail.magehandbook.com (173-8-4-45-WashingtonDC.hfc.comcastbusiness.net [173.8.4.45]) by mx1.freebsd.org (Postfix) with ESMTP id D0EF88FC0C for ; Sun, 10 Apr 2011 14:06:20 +0000 (UTC) Received: from [192.168.1.50] (Mac-Pro.magehandbook.com [192.168.1.50]) by mail.magehandbook.com (Postfix) with ESMTP id 2127D2845B; Sun, 10 Apr 2011 10:06:19 -0400 (EDT) Date: Sun, 10 Apr 2011 10:06:18 -0400 From: Daniel Staal To: Chris Telting , freebsd-questions@freebsd.org Message-ID: <9456D36246CECA618F452F78@Mac-Pro.local> In-Reply-To: <4DA131BD.3030805@telting.org> References: <4DA131BD.3030805@telting.org> X-Mailer: Mulberry/4.0.8 (Mac OS X) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Content-Disposition: inline Cc: Subject: Re: ZFS Striping and Optimizing Capabilities X-BeenThere: freebsd-questions@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list Reply-To: FreeBSD Questions List-Id: User questions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sun, 10 Apr 2011 14:06:21 -0000 --As of April 9, 2011 9:27:41 PM -0700, Chris Telting is alleged to have said: > Does ZFS in any way do performance testing of read/right operating in > light of where the data is stored on the drive? i.e. the outside sectors > of hard drives perform faster. If it does do read/write location testing > can it be shut off or does it detect SSDs? What about tracing > application sector reading and reordering sectors so that they follow one > another according to typical usage? i.e. the sectors are already in the > linear read ahead buffer? --As for the rest, it is mine. I'll let you see Dan Nelson's answer for the striping questions. While I have no inside knowledge of the performance testing handling of ZFS, it doesn't appear to do anything too automagically in that arena. You give it a drive, it will use it, like any other file system. It can give you stats on I/O for each drive as well as the pool in general, and it appears to try to balance reads/writes across all the drives, but that's implied in mirroring or RAID setups. (And remember: ZFS can do either one. You can't add to a RAID volume though, like you can to a mirrored volume.) It does try to cache files, in a fairly aggressive fashion. (By the docs, ZFS will try to fill *all unused RAM* with cached files by default.) And while it doesn't appear to have anything that auto-detects faster drives, you can specify a drive for caching or for the write log. (Caching speeds up reads, and deduplication in versions that support that. The write log speeds up writes.) Telling it to use a SSD for that will speed up those operations considerably. (Note that while a cache drive is considered expendable, the write-log drive is not, and it's recommended that you set up a mirror for it.) Daniel T. Staal --------------------------------------------------------------- This email copyright the author. Unless otherwise noted, you are expressly allowed to retransmit, quote, or otherwise use the contents for non-commercial purposes. This copyright will expire 5 years after the author's death, or in 30 years, whichever is longer, unless such a period is in excess of local copyright law. ---------------------------------------------------------------