From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-qa0-f46.google.com ([209.85.216.46]:62156 "EHLO mail-qa0-f46.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752523Ab3HZR0m convert rfc822-to-8bit (ORCPT ); Mon, 26 Aug 2013 13:26:42 -0400 Received: by mail-qa0-f46.google.com with SMTP id i13so667423qae.12 for ; Mon, 26 Aug 2013 10:26:41 -0700 (PDT) Content-Type: text/plain; charset=us-ascii Mime-Version: 1.0 (Mac OS X Mail 6.5 \(1508\)) Subject: Re: Question: How can I recover this partition? (unable to find logical $hugenum len 4096) From: Nicholas Lee In-Reply-To: <6A12FF1B-5E1A-4F6F-92DA-41E52152E6F2@nickle.es> Date: Mon, 26 Aug 2013 13:26:40 -0400 Cc: linux-btrfs Message-Id: References: <79471CD1-CDDD-4EDD-B255-40568B8446E2@nickle.es> <5EB2ECAC-9A8C-4403-8630-944B646DE3B8@nickle.es> <6A12FF1B-5E1A-4F6F-92DA-41E52152E6F2@nickle.es> To: Chris Murphy Sender: linux-btrfs-owner@vger.kernel.org List-ID: Good news everyone! I got this to mount again using btrfs chunk-recover on the partition! It took around a day for it to run the initial chunk recovery scan (iotop showed that I was limited by disk throughput) and another three hours for the actual chunk recovery (iops limited). Hopefully someone out there that runs into a similar issue will end up finding this! On 22.08.2013, at 21:36, Nicholas Lee wrote: > 6. I don't know off the top of my head, but I'd assume it was the default. It was created by Ubuntu Server Edition > 7. I installed btrfs-progs-git from the AUR, and it (pretty much instantly) returns the following error: > > sudo btrfsck /dev/main-storage-vg/root > Couldn't map the block 1781900460032 > btrfsck: volumes.c:1020: btrfs_num_copies: Assertion `!(ce->start > logical || ce->start + ce->size < logical)' failed. > > If chunk-recover finds anything, I'll let you guys know. (It's still running, and I have yet to write anything.) > > > On 22.08.2013, at 20:58, Chris Murphy wrote: >> >> 6. What was the mkfs.btrfs command used? In particular are you certain the metadata profile is default (DUP)? >> >> 7. If you have a very recent btrfs-progs (few months at most), or better if you can build from btrfs-next, it may be worth running btrfsck *without* the repair option, to see what it has to say about the situation. >