From mboxrd@z Thu Jan 1 00:00:00 1970 From: Dan Mick Subject: Re: rbd STDIN import does not work / wip-rbd-export-stdout Date: Sat, 01 Dec 2012 16:27:29 -0800 Message-ID: <50BAA071.70300@inktank.com> References: <50B36C9F.80107@profihost.ag> <50B36ED1.1020108@filoo.de> <50B37027.5000701@profihost.ag> <2EC6C33E-4909-4AE5-B8C0-5D7A5856EA57@profihost.ag> <50B3CA6F.6000000@inktank.com> <50B484FF.9070500@profihost.ag> <50B62580.2030301@profihost.ag> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Return-path: Received: from mail-da0-f46.google.com ([209.85.210.46]:58674 "EHLO mail-da0-f46.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753809Ab2LBA1t (ORCPT ); Sat, 1 Dec 2012 19:27:49 -0500 Received: by mail-da0-f46.google.com with SMTP id p5so772676dak.19 for ; Sat, 01 Dec 2012 16:27:48 -0800 (PST) In-Reply-To: <50B62580.2030301@profihost.ag> Sender: ceph-devel-owner@vger.kernel.org List-ID: To: Stefan Priebe - Profihost AG Cc: Oliver Francke , "ceph-devel@vger.kernel.org" On 11/28/2012 6:53 AM, Stefan Priebe - Profihost AG wrote: > Hi Dan, > > thanks for your update / new import-export branch. > > The import seems to work now. What i'm wondering is that it is really > slow. > > I see around 10% CPU Usage on importing host (host where rbd import is > started) and just 4MB every 2s raising Data on ceph. This means 40GB > took around 5-6 hours. > > Night this be due to he size update? 4MB every 2s looks like the > stripe size. I did change the resize to double each time to help with that before pushing to master, but that still seems awfully slow. You should definitely use the version in the mainline, though; it's better in several ways. Are these particularly sparse images? I'm testing a "preserve sparseness" tweak that should help a lot if so (and save space in the cluster, obviously).