From: Jan Kara Subject: Re: [PATCH 1/2] [PATCH] ext4: Add inode to the orphan list during block allocation failure Date: Sat, 6 Jun 2009 01:42:39 +0200 Message-ID: <20090605234238.GF11650@duck.suse.cz> References: <20090331044544.GB5979@skywalker> <1238491766-13182-1-git-send-email-aneesh.kumar@linux.vnet.ibm.com> <20090405031116.GG7553@mit.edu> <20090406100509.GB31189@duck.suse.cz> <20090605043117.GB4046@mit.edu> <20090605062234.GA27217@mit.edu> <20090605072417.GA8236@mit.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: "Aneesh Kumar K.V" , linux-ext4@vger.kernel.org To: Theodore Tso Return-path: Received: from cantor2.suse.de ([195.135.220.15]:33132 "EHLO mx2.suse.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1756520AbZFEXmj (ORCPT ); Fri, 5 Jun 2009 19:42:39 -0400 Content-Disposition: inline In-Reply-To: <20090605072417.GA8236@mit.edu> Sender: linux-ext4-owner@vger.kernel.org List-ID: On Fri 05-06-09 03:24:17, Theodore Tso wrote: > On Fri, Jun 05, 2009 at 02:22:34AM -0400, Theodore Tso wrote: > > I've tried applying these two patches to the ext4 patch queue. On a > > metadata-heavy workload (specifically, fsx), it causes a 5% > > degradation in the wall clock run-time of the fsx run-time. > > I just did some more timing tests, and it looks like I can't confirm > it. The test times are noisy enough I need to run some more under very > controlled circumsntances. So ignore this for now, it might not be a > problem after all. It would be strange if the patches caused a measurable slowdown. They change something only in the failure path where we fail to allocate some blocks or fail to copy in all the data and these should better be exceptional cases... Honza -- Jan Kara SUSE Labs, CR