From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:34956) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1b8qpv-0002cj-Va for qemu-devel@nongnu.org; Fri, 03 Jun 2016 11:16:42 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1b8qps-0007hn-2F for qemu-devel@nongnu.org; Fri, 03 Jun 2016 11:16:35 -0400 Date: Fri, 3 Jun 2016 17:16:21 +0200 From: Kevin Wolf Message-ID: <20160603151621.GC5712@noname.redhat.com> References: <1464957021-3469-1-git-send-email-kwolf@redhat.com> <1464957021-3469-6-git-send-email-kwolf@redhat.com> <57519BBE.5040204@openvz.org> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <57519BBE.5040204@openvz.org> Subject: Re: [Qemu-devel] [PATCH 5/5] qemu-img bench: Add --flush-interval List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: "Denis V. Lunev" Cc: qemu-block@nongnu.org, mreitz@redhat.com, eblake@redhat.com, qemu-devel@nongnu.org Am 03.06.2016 um 17:01 hat Denis V. Lunev geschrieben: > On 06/03/2016 03:30 PM, Kevin Wolf wrote: > >This options allows to flush the image periodically during write tests. > > > >Signed-off-by: Kevin Wolf > This pattern could be a bit different - you wait requests to finish > and after that start flush. In this case there is no gain with my > patch ;) > > You should treat flush as ordinary write request i.e. place > request immediately without waiting for writes to finish. > May be this could be specified as an operation mode. Hm, okay, I can try that. What OS is running on your guest? Because I seem to remember that Linux drains the queue before sending a flush command. This is why I implemented it this way here. Maybe it changed in Linux, though. Kevin