We recently replaced a hard disk in a server and reinstalled the operating system (Centos 6)
The raid rebuilds itself every boot; /dev/md2 is losing a disk each time apparently. Always the same array, /dev/md2 :
# mdadm --detail /dev/md2
/dev/md2:
Version : 1.0
Creation Time : Tue Jul 31 19:26:14 2012
Raid Level : raid1
Array Size : 1462516600 (1394.76 GiB 1497.62 GB)
Used Dev Size : 1462516600 (1394.76 GiB 1497.62 GB)
Raid Devices : 2
Total Devices : 2
Persistence : Superblock is persistent
Update Time : Mon Aug 20 16:07:51 2012
State : active, resyncing
Active Devices : 2
Working Devices : 2
Failed Devices : 0
Spare Devices : 0
Resync Status : 13% complete
Name : rescue:2
UUID : dfdcd9c4:24381dd7:25ffb77a:9bc2784c
Events : 112
Number Major Minor RaidDevice State
0 8 3 0 active sync /dev/sda3
1 8 19 1 active sync /dev/sdb3
The only inconsistency I can find (I am no expert at RAID) is that blkid comes up with the following:
# blkid
/dev/sda3: UUID="73315a4d-2885-45ed-88e9-00c66d449115" TYPE="ext4"
/dev/sdb3: UUID="dfdcd9c4-2438-1dd7-25ff-b77a9bc2784c" UUID_SUB="72ffe87d-3105-dd6c-5b5c-58b14c2afc7f" LABEL="rescue:2" TYPE="linux_raid_member"
/dev/md2: UUID="73315a4d-2885-45ed-88e9-00c66d449115" TYPE="ext4"
The UUID reported by blkid for md2 is not the same as the one reported by mdadm. And also, sda3 has the same UUID in blkid as md2.
Best Answer
Hehe six months later... but I'll try to answer nonetheless.
mdadm --examine --scan
shows you the several RAIDs' uuids which should be consistent withmdadm.conf
For each RAID, all the partitions that make the RAID share this same uuid.blkid
gives you the filesystem uuids which should be consistent with fstabSo
blkid
should show you something like:where
/dev/sd[ab]3
's uuid is the RAID uuidYou could try to recreate the RAID.
Is
/dev/sda3
part of a RAID? (You can check that withmdadm --examine /dev/sda3
)If it is, then you can do something similar to this: