From mboxrd@z Thu Jan 1 00:00:00 1970 From: Neil Brown Subject: Re: Request on RAID10 Date: Tue, 29 Jun 2010 16:53:27 +1000 Message-ID: <20100629165327.0a05a619@notabene.brown> 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: koti Cc: linux-raid@vger.kernel.org List-Id: linux-raid.ids On Sat, 26 Jun 2010 16:56:33 +0000 (UTC) koti wrote: > Hi Everyone, > > We created RAID10 array on 4 disks (physically we had 3 1TB disks as test case > we created 350GB partitions). We got a hang while we are doing the following: > 1) remove the disk and add this again. Doing this in while loop. > 2) While doing the above, simultaneously read and write on the same array. > > At some point of time the recovery going to hang, then the array was not > operational, we are using kernel 2.6.27. > > I am new to use the RAID 10 array. Is this scenario was correct? if it is then > what might be the problem, can any body suggest me and direct in proper way. It > is very much useful for proceed further on this. Hi Koti, it looks like you might have found and interesting bug. Some sort of race around shutdown or startup. It would be very helpful if you could post some details about the exact situation when the system has hung. cat /proc/mdstat would be helpful (if that doesn't hang), and echo t > /proc/sysrq-trigger then collect the trace backs that went to the kernel logs and post them. Thanks, NeilBrown