mdadm concerns

tomplancon7092 · 2017-05-04T18:44:50.000Z

Hi all,

I have a CentOS 6.8 server that is randomly restarting and throwing memory errors. I’m getting ready to move the data disks to another CentOS server. The disks on the failing server are in a md raid 1 array. I looked at the mdadm.conf file and saw:

ARRAY /dev/md2 metadata=1.1 name=localhost.localdomain:2 UUID=57df03e7:b056b5bd:93929c7c:46277171
ARRAY /dev/md1 metadata=1.1 name=localhost.localdomain:1 UUID=9ad31155:a738c7ef:a8c11c19:e341fb5c
ARRAY /dev/md3 metadata=1.1 name=localhost.localdomain:3 UUID=8f61067e:dd5843a9:121bfb9d:89a7db9c
ARRAY /dev/md0 metadata=1.0 name=localhost.localdomain:0 UUID=b1237535:763092f5:c8e7f62c:b139b8f6
ARRAY /dev/md/localhost.localdomain:4 metadata=1.1 name=localhost.localdomain:4 UUID=1cdd508e:c5435144:af0b3014:04ea2681

That did not seem correct, so I checked /proc/mdstat:

Personalities : [raid1] 
md0 : active raid1 sda1[0] sdb1[1]
      102388 blocks super 1.0 [2/2] [UU]
      
md3 : active raid1 sda5[0] sdb5[1]
      264258428 blocks super 1.1 [2/2] [UU]
      bitmap: 0/2 pages [0KB], 65536KB chunk

md127 : active raid1 sdc[0] sdd[1]
      976631488 blocks super 1.2 [2/2] [UU]
      bitmap: 4/8 pages [16KB], 65536KB chunk

md1 : active raid1 sdb2[1] sda2[0]
      20478908 blocks super 1.1 [2/2] [UU]
      bitmap: 1/1 pages [4KB], 65536KB chunk

md2 : active raid1 sdb3[1]
      8190968 blocks super 1.1 [2/1] [_U]
      
unused devices: <none>

The md127 is the array holding all the data so I was surprised not to see it referenced in mdadm. conf. So I did a mdadm --detail --scan >> /etc/mdadm.conf and saw this in the updated mdadm.conf:

ARRAY /dev/md2 metadata=1.1 name=localhost.localdomain:2 UUID=57df03e7:b056b5bd:93929c7c:46277171
ARRAY /dev/md2 metadata=1.1 name=localhost.localdomain:2 UUID=57df03e7:b056b5bd:93929c7c:46277171
ARRAY /dev/md1 metadata=1.1 name=localhost.localdomain:1 UUID=9ad31155:a738c7ef:a8c11c19:e341fb5c
ARRAY /dev/md1 metadata=1.1 name=localhost.localdomain:1 UUID=9ad31155:a738c7ef:a8c11c19:e341fb5c
ARRAY /dev/md3 metadata=1.1 name=localhost.localdomain:3 UUID=8f61067e:dd5843a9:121bfb9d:89a7db9c
ARRAY /dev/md3 metadata=1.1 name=localhost.localdomain:3 UUID=8f61067e:dd5843a9:121bfb9d:89a7db9c
ARRAY /dev/md0 metadata=1.0 name=localhost.localdomain:0 UUID=b1237535:763092f5:c8e7f62c:b139b8f6
ARRAY /dev/md0 metadata=1.0 name=localhost.localdomain:0 UUID=b1237535:763092f5:c8e7f62c:b139b8f6
ARRAY /dev/md/localhost.localdomain:4 metadata=1.1 name=localhost.localdomain:4 UUID=1cdd508e:c5435144:af0b3014:04ea2681
ARRAY /dev/md/pequod.domus-bka.local:5 metadata=1.2 name=pequod.domus-bka.local:5 UUID=dfb4867a:7b1c4e11:3ee7ac47:67dfb2e7

Aside from the obvious duplication, why am I not seeing a reference to /dev/md127 ? And are the last 2 lines actually the md127 device? If so, why are they different?

Any ifo here is greatly appreciated. I’d really like to be able to move this array intact if possible!

Thanks much!

alexgaft · 2017-05-07T19:22:23.000Z

Maybe because of metadata version?

tomplancon7092 · 2017-05-08T12:24:07.000Z

Thanks for the reply.

Sorry, I’m not sure what that means, or at least the implications of it.

woland · 2017-05-09T06:51:01.000Z

Haven’t played with software raid much, and it was a while back. What does your fstab look like?

What about the blkid lsblk and finally df -h commands

md0 appears to be a boot volume perhaps?

md2 might be swap but looks like it’s degraded, sda3 missing?

The other three are probably root and data partitions?