On Sun, Jan 3, 2016 at 9:16 PM, NeilBrown <neilb@xxxxxxxx> wrote:
> On Sun, Dec 27 2015, George Rapp wrote:
>
> If you apply the following patch to mdadm (
> git clone git://neil.brown.name/mdadm
> apply patch
> make
> ) and then try to assemble with --update=revert-reshape, it should
> assemble as a 5-device array with no reshape happening.
>
> I probably want more safety checks before this goes upstream, but it
> is safe enough for you.
>
> [patch snipped]
>
I've replied privately to Neil, but for list archive purposes, his
patch solved my problem.
After applying his patch, I used the custom-build version of mdadm to
assemble my array:
# UUID=$(mdadm -E /dev/sdd4 | perl -ne '/Array UUID : (\S+)/ and print $1')
# DEVICES=$(cat /proc/partitions | parallel --tagstring {5} --colsep '
+' mdadm -E /dev/{5} |grep $UUID | parallel --colsep '\t' echo
/dev/{1})
# ./mdadm --assemble --verbose --update=revert-reshape /dev/md4 $DEVICES
mdadm: looking for devices for /dev/md4
mdadm: /dev/sdd4 is identified as a member of /dev/md4, slot 0.
mdadm: /dev/sdc4 is identified as a member of /dev/md4, slot 1.
mdadm: /dev/sdg4 is identified as a member of /dev/md4, slot 2.
mdadm: /dev/sdh1 is identified as a member of /dev/md4, slot 3.
mdadm: /dev/sdi1 is identified as a member of /dev/md4, slot 5.
mdadm: /dev/sdj1 is identified as a member of /dev/md4, slot 4.
mdadm: device 10 in /dev/md4 has wrong state in superblock, but
/dev/sdi1 seems ok
mdadm: added /dev/sdc4 to /dev/md4 as 1
mdadm: added /dev/sdg4 to /dev/md4 as 2
mdadm: added /dev/sdh1 to /dev/md4 as 3
mdadm: added /dev/sdj1 to /dev/md4 as 4
mdadm: added /dev/sdi1 to /dev/md4 as 5
mdadm: added /dev/sdd4 to /dev/md4 as 0
mdadm: /dev/md4 has been started with 5 drives and 1 spare.
# cat /proc/mdstat
Personalities : [raid1] [raid6] [raid5] [raid4]
md4 : active raid6 sdd4[0] sdi1[8](S) sdj1[7] sdh1[6] sdg4[5] sdc4[1]
5756723712 blocks super 1.2 level 6, 512k chunk, algorithm 2 [5/5] [UUUUU]
bitmap: 0/15 pages [0KB], 65536KB chunk
For future reference, I would highly recommend executing
# setenforce 0
or, if you're really paranoid, disabling SELinux completely
- edit /etc/sysconfig/selinux and add or modify this line:
SELINUX=disabled
- # touch /.relabel
- # systemctl reboot
(and go get a cup of coffee while your filesystem gets relabeled)
before making any metadata changes on a RAID 6 array.
Thanks again, Neil and members of the Linux-RAID community!
--
George Rapp (Pataskala, OH) Home: george.rapp -- at -- gmail.com
LinkedIn profile: https://www.linkedin.com/in/georgerapp
Phone: +1 740 936 RAPP (740 936 7277)
--
To unsubscribe from this list: send the line "unsubscribe linux-raid" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
