Hello Phil,

You are 100% correct. Thanks for your email. I had 2 separate ARRAY 
entries in my mdadm.config file. When I added the spare, I added the new 
information to the config file but did not remove the old entry. Total 
newbie mistake.

- I commented out all the previous entries in the mdadm.config file (it 
had a devices lists instead of the UUID information)

- I successfully assembled the array (mdadm --assemble /dev/md0  
/dev/sdb1 etc...)

- I  updated the config file using the information from mdadm -Esc 
partitions | grep UUID and deleted the old entries.

- I uncommented the fstab entry for the array and rebooted.


Everything is working perfectly now.


I remember reading somewhere that mdadm can run without the config file 
by using the RAID uuid and the information in the superblocks. Would 
this be more reliable?

Thanks Again,

- Justin

------ Original Message ------
From: "Phil Turmel" <redacted>
To: "Justin Stephenson" <redacted>; 
linux-raid@vger.kernel.org
Sent: 10/02/2014 12:53:31 PM
Subject: Re: Reassemble Raid 6

On 02/10/2014 11:48 AM, Justin Stephenson wrote:

quoted

 Hello,

 I have reviewed the Linux Raid Wiki regarding rebuilding and it
 recommended writing a note to the list.

 I am running an mdadm raid on CENTOS 6.5 with 6 disks in a Raid 6 off 
of
 my motherboard's SATA controller.

 I added a 7th disk as a spare.

 When I rebooted the raid after adding the spare, it would not mount 
and
 the OS would only boot into recovery mode. I changed my fsck file to
 exclude the raid and can now boot into the GUI.

 I have tried to reassemble the array using --assemble and I get the 
the
 following message:

 mdadm: we match both /dev/md/BigBlue:0 and /dev/md0 - cannot decide
 which to use.

Please show the output of your mdadm.conf file, and the complete 
command
you used to assemble with. Add "-v" to it to show more information.

{There's nothing wrong with your array, just conflicting info in your
configuration file.}

Phil