From mboxrd@z Thu Jan 1 00:00:00 1970 From: Neil Brown Subject: Re: RAID10 status when you remove the first disk and last disk Date: Wed, 21 Jul 2010 13:06:50 +1000 Message-ID: <20100721130650.3b23b960@notabene> References: Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: Sender: linux-raid-owner@vger.kernel.org To: Michael Li Cc: linux-raid@vger.kernel.org List-Id: linux-raid.ids On Tue, 20 Jul 2010 19:34:06 -0700 "Michael Li" wrote: > Hi , > > > > I am using RAID10 far copy 2 by 4 disks, when I removed the first disk > and last disk, I think this raid should be failed, but mdadm -D > /dev/md2, I saw it was degraded, but when I try to read/write, there > will be I/O error. And when I added the two disks again, the raid will > be recovery, so that's normal? why it is designed like this? > > > > Anyone know the detail about that? Thanks so much! > Sounds like a bug. You have definitely lost data if the first and last devices of such an array go missing, so it should not try to recover a spare in to either slot. What kernel version are you using? NeilBrown