From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Yan, Zheng" Subject: Re: [PATCH 00/39] fixes for MDS cluster recovery Date: Mon, 01 Apr 2013 16:46:37 +0800 Message-ID: <5159496D.6040002@intel.com> References: <1363531902-24909-1-git-send-email-zheng.z.yan@intel.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Return-path: Received: from mga02.intel.com ([134.134.136.20]:31989 "EHLO mga02.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1756021Ab3DAIqu (ORCPT ); Mon, 1 Apr 2013 04:46:50 -0400 In-Reply-To: <1363531902-24909-1-git-send-email-zheng.z.yan@intel.com> Sender: ceph-devel-owner@vger.kernel.org List-ID: To: ceph-devel@vger.kernel.org, sage@inktank.com, greg@inktank.com On 03/17/2013 10:51 PM, Yan, Zheng wrote: > From: "Yan, Zheng" > > This serie fixes issues I encountered when running random MDS restart tests. > With these patches, my 3 MDS setup that runs fsstress + thrash_exports can > survive restarting one or two MDS dozens of times. > > But there still are lots of unsolved problems, Sometimes rstat corruption, > request hangs ... > > This patch series are also in: > git://github.com/ukernel/ceph.git wip-mds > I rebased these patches. Replaced patch 5 and patch 6 with two new patches. [PATCH] mds: avoid sending duplicated table prepare/commit [PATCH] mds: don't roll back prepared table updates Regards Yan, Zheng