From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from merlin.infradead.org ([205.233.59.134]:36940 "EHLO merlin.infradead.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1757166Ab3E0IWP (ORCPT ); Mon, 27 May 2013 04:22:15 -0400 Date: Mon, 27 May 2013 10:21:47 +0200 From: Jens Axboe Subject: Re: Question on --buffer_compress_percentage Message-ID: <20130527082147.GC29680@kernel.dk> References: MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: Sender: fio-owner@vger.kernel.org List-Id: fio@vger.kernel.org To: Matthew Eaton Cc: fio@vger.kernel.org On Fri, May 24 2013, Matthew Eaton wrote: > Hello fio group, > > Can anyone tell me if --buffer_compress_percentage fills the buffer > with random data and reuses that data or if it refills with random > data for each submit? If you want the latter behavior should you > combine with --refill_buffers? That is actually explained in the man page / HOWTO: buffer_compress_percentage=int If this is set, then fio will attempt to provide IO buffer content (on WRITEs) that compress to the specified level. Fio does this by providing a mix of random data and zeroes. Note that this is per block size unit, for file/disk wide compression level that matches this setting, you'll also want to set refill_buffers. So the answer to your question is yes. Fio will reuse IO buffers, unless --refill_buffers is set. If you do not set refill_buffers with buffer_compress_percentage, then you will have identically compressible blocks written. And that would only give you the per-block compression ratio specified, on a broader scale it would compress more. -- Jens Axboe