From owner-freebsd-questions@freebsd.org Fri Jan 22 05:05:30 2016 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 0E328A8C075 for ; Fri, 22 Jan 2016 05:05:30 +0000 (UTC) (envelope-from mason@blisses.org) Received: from phlegethon.blisses.org (phlegethon.blisses.org [50.56.97.101]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (Client did not present a certificate) by mx1.freebsd.org (Postfix) with ESMTPS id E88A11542 for ; Fri, 22 Jan 2016 05:05:29 +0000 (UTC) (envelope-from mason@blisses.org) Received: from blisses.org (cocytus.blisses.org [23.25.209.73]) by phlegethon.blisses.org (Postfix) with ESMTPSA id AB7401493BE; Fri, 22 Jan 2016 00:05:27 -0500 (EST) Date: Fri, 22 Jan 2016 00:05:26 -0500 From: Mason Loring Bliss To: "Michael B. Eichorn" Cc: dweimer , freebsd-questions@freebsd.org Subject: Re: ZFS performance help sought Message-ID: <20160122050526.GO4538@blisses.org> References: <20160121205139.GG4538@blisses.org> <20160121232810.GJ4538@blisses.org> <20160121235545.GK4538@blisses.org> <20160122000523.GL4538@blisses.org> <1453438096.26786.7.camel@michaeleichorn.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <1453438096.26786.7.camel@michaeleichorn.com> User-Agent: Mutt/1.5.23 (2014-03-12) 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, 22 Jan 2016 05:05:30 -0000 On Thu, Jan 21, 2016 at 11:48:16PM -0500, Michael B. Eichorn wrote: > freebsd.org/doc/en_US.ISO8859-1/books/handbook/zfs-advanced.html Hm. I don't know that I've looked at this before. Thank you. I'll read it. > I suppose vfs.zfs.top_maxinflight might worth a try? Sure. Does it have any correspondence with the qlen iostat -x reports? I don't know how many levels have queueing, and whether iostat is reading the drive's tagged command queueing, something in the filesystem, or something elsewhere in the kernel. I'll try to figure this out tomorrow, as well as reading the rest of that page and testing with top_maxinflight. If iostat is reporting something else, is there a good way to capture current inflight to see how close to the cap I'm getting in any event? -- Mason Loring Bliss mason@blisses.org http://blisses.org/ "I am a brother of jackals, and a companion of ostriches." (Job 30 : 29)