From mboxrd@z Thu Jan 1 00:00:00 1970 From: Hullen@t-online.de (Helmut Hullen) Subject: Re: kernel 3.3.4 damages filesystem (?) Date: 08 May 2012 22:19:00 +0200 Message-ID: References: <20120508200228.GM8938@carfax.org.uk> Reply-To: helmut@hullen.de Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII To: linux-btrfs@vger.kernel.org Return-path: In-Reply-To: <20120508200228.GM8938@carfax.org.uk> List-ID: Hallo, Hugo, Du meintest am 08.05.12: >>> Otherwise if you remove a disk from a raid0 (doesn't matter if you >>> have 2 or 5 or x disks in the fs, btrfs should stripe above all >>> disks) your fs should be broken. >> Not with btrfs ... there it works even with >> >> mkfs.btrfs -m raid1 -d raid0 ... > There is a big difference between "orderly and planned removal of > a hard disk", and "disk goes away with no warning". And I know the difference ... When I first called for help I searched the failure in another place than in "disk is dead". > This is essentially the difference you've been talking about at cross- > purposes all day. What I still hope (may be it's impossible): when 1 disk/partition fails, then the contents of the other disks is "somehow" restorable. And not irreproducable. Viele Gruesse! Helmut