From mboxrd@z Thu Jan 1 00:00:00 1970 From: Adam Thompson Subject: Re: [Board] dead RAID6 array on CentOS6.6 / kernel 3.19 Date: Tue, 10 Feb 2015 21:37:36 -0600 Message-ID: <54DACE80.3010004@athompso.net> References: Reply-To: athompso@athompso.net Mime-Version: 1.0 Content-Type: text/plain; charset=windows-1252; format=flowed Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: Sender: linux-raid-owner@vger.kernel.org To: Trevor Cordes Cc: linux-raid@vger.kernel.org List-Id: linux-raid.ids On 2015-02-10 08:58 PM, Trevor Cordes wrote: > The big first question is was it lying when it said starting bg recon. > Probably. How can it rebuild a 3+ drive failure. > > If so it should be possible to really force. As in wild wacky stuff. > I've done that before. > > I can look at it more after the meeting our tomorrow morning. > > Was the reboot a clean or a sysreq? Did it sync first? Manually initiated by reboot(8), which is a link to telinit(8), which merely sends a signal to init(8), which actually did the reboot. No idea if /dev/md127 correctly stopped or not, syslogd stops before those messages get logged, and I was working remotely through SSH instead of to the virtual console connection. I would assume the sync(2) happened correctly; this is breakage way down at the MD layer, not the FS layer. You're right, RAID6 can't rebuild from a 3-drive failure. I've gone back through the messages I sent, and it looks like there was no hot-spare, so I suspect this is just the default code path, where it tries to rebuild anything it can, then figures out it can't. -- -Adam Thompson athompso@athompso.net +1 (204) 291-7950 - cell +1 (204) 489-6515 - fax