From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751875Ab0CTF6o (ORCPT ); Sat, 20 Mar 2010 01:58:44 -0400 Received: from mail-gy0-f174.google.com ([209.85.160.174]:52976 "EHLO mail-gy0-f174.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751472Ab0CTF6n convert rfc822-to-8bit (ORCPT ); Sat, 20 Mar 2010 01:58:43 -0400 DomainKey-Signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=subject:mime-version:content-type:from:in-reply-to:date:cc :content-transfer-encoding:message-id:references:to:x-mailer; b=jP5XJsVAquiWOEzHD8eSU4NsYRkXoO8K4hERrNYRVHWcfJKiwlI05jCV8GF1hZFVs8 q7lfoK65qVuFu5MTTMdMwNcC2vj5PLq7ScPBds/wmJ8c9idb0J3i0rDE+klDGbiGXYb6 eS7m9wWHgzch3P5/hEAmFsmfTN8rpV93kmoUc= Subject: Re: Balance taking a lot longer than before Mime-Version: 1.0 (Apple Message framework v1077) Content-Type: text/plain; charset=windows-1252 From: Mark Rada In-Reply-To: Date: Sat, 20 Mar 2010 01:58:39 -0400 Cc: Zheng Yan Content-Transfer-Encoding: 8BIT Message-Id: <0FB2627B-B950-4062-80D5-0BAEC3AF7AB3@gmail.com> References: <77201DCC-CC67-4B00-8FFE-AAB17A5D18DA@mailservices.uwaterloo.ca> <3d0408631003182019i1509d363j98f94661679ea30b@mail.gmail.com> <3d0408631003190032s663624c1o528674f8010309f2@mail.gmail.com> To: linux-kernel@vger.kernel.org X-Mailer: Apple Mail (2.1077) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org I'm happy to report that defragmenting fixed my problem. In fact, it did a really good job; I had to wrap it to recurse through a few folders I suspected were highly fragmented which took about half an hour run it a few times. Then I tried to rebalance again and not only did it work, but it only took a few hours. — Mark Rada markrada26@gmail.com On 2010-03-19, at 10:18 AM, Mark Rada wrote: > I don't have any files bigger than 5GB, and I'm using less than a quarter of my total disk space but I already have a lot of fragmentation? > > Defragmenting won't help, will it? Is there anything I can do other than deleting the files? > > — > Mark Rada > markrada26@gmail.com > > > On 2010-03-19, at 3:32 AM, Yan, Zheng wrote: > >> The loop is caused by fragmentaion of free space. Balance code is trying to >> relocate a large file extent, but it can't find a free space that is >> large enough. >> >> On Fri, Mar 19, 2010 at 11:28 AM, Mark Rada wrote: >>> Hmm, there seems to be a very very long list of "btrfs: found 1 extents" messages and nothing else in between them. >>> >>> Why would deleting large files help? I don't want to delete files on my server unless I have no other option. >>> >>> — >>> Mark Rada >>> marada@uwaterloo.ca >>> >>> >>> On 2010-03-18, at 11:19 PM, Yan, Zheng wrote: >>> >>>> On Fri, Mar 19, 2010 at 9:44 AM, Mark Rada wrote: >>>>> Hi devs, >>>>> >>>>> I've been using btrfs on my file server since 2.6.32 and after upgrading to .33 I noticed that rebalancing took way longer than before. >>>>> >>>>> I have a 5 disk array used for a btrfs md setup, different hard drive sizes, brands, and interfaces. >>>>> >>>>> The last time I rebalanced I was using about 800GB of space on the 4.6TB array, and I started before I went to sleep and it was done when I woke up. >>>>> >>>>> I'm in the middle of a rebalance right now, currently using 950GB, and it has been a full day and the rebalancing is not done yet. Is this an expected change? Is there any way to tell how much longer it will take? Do you need to know more info about the setup? >>>>> >>>> >>>> I guess it enters infinite loops. you can check it by checking if >>>> there are a lots of >>>> repeats messages in the kernel log. If it is true, you can try >>>> deleting some large >>>> files in that FS. >>> >