From mboxrd@z Thu Jan 1 00:00:00 1970 From: Josh Durgin Subject: Re: Poor read performance in KVM Date: Wed, 18 Jul 2012 08:34:43 -0700 Message-ID: <5006D793.6090906@inktank.com> References: <5002C215.108@bashkirtsev.com> <5003B1CC.4060909@inktank.com> <50064DCD.8040904@bashkirtsev.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Return-path: Received: from mail-gh0-f174.google.com ([209.85.160.174]:33643 "EHLO mail-gh0-f174.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751800Ab2GRPeq (ORCPT ); Wed, 18 Jul 2012 11:34:46 -0400 Received: by ghrr11 with SMTP id r11so1686884ghr.19 for ; Wed, 18 Jul 2012 08:34:45 -0700 (PDT) In-Reply-To: <50064DCD.8040904@bashkirtsev.com> Sender: ceph-devel-owner@vger.kernel.org List-ID: To: Vladimir Bashkirtsev Cc: ceph-devel@vger.kernel.org On 07/17/2012 10:46 PM, Vladimir Bashkirtsev wrote: >> ceph --admin-daemon /var/run/ceph/kvm.asok perf dump > Done that. Waited for VM to fully boot then got perf dump. It would be > nice to get output in human readable format instead of JSON - I remember > some other part of ceph had relevant command line switch. Does it exist > for perf dump? I forgot to mention you can pipe that to 'python -mjson.tool' for more readable output. It's intended to be used by monitoring tools, hence json, but some kind of more plain output could be added. Josh