From mboxrd@z Thu Jan 1 00:00:00 1970 From: Robert Schultz Subject: Re: RAID 5 3-drive array failed 2 disks at once - can anything be saved? Date: Wed, 18 Sep 2013 22:29:51 -0400 Message-ID: <523A619F.90302@schultzfamily.ca> References: <52332763.30901@schultzfamily.ca> <52347194.7010602@turmel.org> <52361B98.9030705@schultzfamily.ca> <52365AE8.5010807@turmel.org> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <52365AE8.5010807@turmel.org> Sender: linux-raid-owner@vger.kernel.org To: linux-raid@vger.kernel.org Cc: Phil Turmel List-Id: linux-raid.ids That all worked beautifully. Right up until I left the BackupPC running against a RAID array with a bad disk. It failed again after about 30 hours. The symptoms are the same. I think I need to bring the array back up but leave that disk offline with: mdadm --assemble --force /dev/md0 /dev/sdb /dev/sdc (sdd is the bad drive) Then follow the remainder of the steps to check. I have a new disk on the way. I would then add this new disk into the array and sync. Does that sound correct? Rob On 13-09-15 09:12 PM, Phil Turmel wrote: > Hi Robert, > > On 09/15/2013 04:42 PM, Robert Schultz wrote: >> Phil: >> >> Thank you for the information. This is my backup machine. Up to this >> point I wasn't concerned about having a second copy of this machine, but >> I have a tendency to decommission a computer and leave the backups on by >> backuppc for archive purposes. I probably don't really, really need >> anything on this PC. That said I'm am very paranoid that I will have >> some other failure before I can resolve this :-( >> >> I hadn't read anything about timing in disks in RAID - I'll have to go >> do some research. I see WD has their RED series that appears to be >> directed to this market. > Please do read the archives on the topic. You won't regret it. > > And yes, the WD REDs power up with SCTERC set properly. I bought four > of these for my new media server. > >> Here is the information requested. Please let me know if this changes >> anything in your instructions. I'll hold off until you confirm. > One modest change. Two of your drives *do* support SCTERC, they just > have to have it enabled on every powerup: > >> SCT Error Recovery Control: >> Read: Disabled >> Write: Disabled > For those two drives, your boot sequence should have: > > smartctl -l scterc,70,70 /dev/sdX > > For the other, you still need: > > echo 180 >/sys/block/sdX/device/timeout > > Otherwise, my recommendations stand. > > Phil