mdadm UUID lost after reboot

1

Here is the problem. I part all disks into 4 partitions. Use all disk partition one to create raid1(md0). However, after reboot some of the raid members lost, the raid is degraded. Use blkid to check the UUID and fount it lost.

Create command:

mdadm -C ${RAID_CHECK} -l 1 -n ${disk_num} ${disk_str};

and then write config:

mdadm --detail --scan > /etc/mdadm.conf

After creating:

mdadm -D /dev/md0 shows:

[root@afa_primary ~]# mdadm -D /dev/md0
/dev/md0:
           Version : 1.2
     Creation Time : Fri Jul 26 13:58:44 2019
        Raid Level : raid1
        Array Size : 1039360 (1015.00 MiB 1064.30 MB)
     Used Dev Size : 1039360 (1015.00 MiB 1064.30 MB)
      Raid Devices : 8
     Total Devices : 8
       Persistence : Superblock is persistent

       Update Time : Mon Jul 29 11:10:54 2019
             State : clean
    Active Devices : 8
   Working Devices : 8
    Failed Devices : 0
     Spare Devices : 0

Consistency Policy : resync

              Name : afa_primary:0  (local to host afa_primary)
              UUID : 5645af3f:a0bc4deb:8f876e72:0520f210
            Events : 149

    Number   Major   Minor   RaidDevice State
       0     259        7        0      active sync   /dev/nvme1n1p1
       1     259        2        1      active sync   /dev/nvme2n1p1
       2     259       23        2      active sync   /dev/nvme3n1p1
       8     259       18        3      active sync   /dev/nvme4n1p1
       9     259       27        4      active sync   /dev/nvme5n1p1
      10     259       38        5      active sync   /dev/nvme6n1p1
       6     259       14        6      active sync   /dev/nvme7n1p1
      11     259       42        7      active sync   /dev/nvme8n1p1

use blkid to check every member:

[root@afa_primary ~]# blkid /dev/nvme4n1p1
/dev/nvme4n1p1: UUID="5645af3f-a0bc-4deb-8f87-6e720520f210" UUID_SUB="82e5c45d-857e-b54b-f4d8-c1f8a83e61a1" LABEL="afa_primary:0" TYPE="linux_raid_member" PARTLABEL="primary" PARTUUID="9cb306a4-527f-47ec-a134-f0272b185108"
You have new mail in /var/spool/mail/root
[root@afa_primary ~]# blkid /dev/nvme5n1p1
/dev/nvme5n1p1: UUID="5645af3f-a0bc-4deb-8f87-6e720520f210" UUID_SUB="bfe17ad0-3209-47c0-54df-853c0e59618d" LABEL="afa_primary:0" TYPE="linux_raid_member" PARTLABEL="primary" PARTUUID="79e1b000-51e3-4742-8f9c-8c0eb9ab29cc"
[root@afa_primary ~]#

My disk list is:

[root@afa_primary ~]# nvme list
Node             SN                   Model                                    Namespace Usage                      Format           FW Rev
---------------- -------------------- ---------------------------------------- --------- -------------------------- ---------------- --------
/dev/nvme0n1     PHHH9012029A128A     INTEL SSDPEKKA128G8                      1         128.04  GB / 128.04  GB    512   B +  0 B   006D
/dev/nvme1n1     FL191900274          P5510DS0384T00                           1           3.84  TB /   3.84  TB    512   B +  0 B   224003K0
/dev/nvme2n1     FL191900379          P5510DS0384T00                           1           3.84  TB /   3.84  TB    512   B +  0 B   224003K0
/dev/nvme3n1     FL191900343          P5510DS0384T00                           1           3.84  TB /   3.84  TB    512   B +  0 B   224003K0
/dev/nvme4n1     FL191900355          P5510DS0384T00                           1           3.84  TB /   3.84  TB    512   B +  0 B   224003K0
/dev/nvme5n1     FL191900222          P5510DS0384T00                           1           3.84  TB /   3.84  TB    512   B +  0 B   224003K0
/dev/nvme6n1     FL191900349          P5510DS0384T00                           1           3.84  TB /   3.84  TB    512   B +  0 B   224003K0
/dev/nvme7n1     FL191900372          P5510DS0384T00                           1           3.84  TB /   3.84  TB    512   B +  0 B   224003K0
/dev/nvme8n1     FL191900214          P5510DS0384T00                           1           3.84  TB /   3.84  TB    512   B +  0 B   224003K0

after reboot,

[root@afa_primary ~]# mdadm -D /dev/md0
/dev/md0:
           Version : 1.2
     Creation Time : Mon Jul 29 11:54:37 2019
        Raid Level : raid1
        Array Size : 1039360 (1015.00 MiB 1064.30 MB)
     Used Dev Size : 1039360 (1015.00 MiB 1064.30 MB)
      Raid Devices : 8
     Total Devices : 4
       Persistence : Superblock is persistent

       Update Time : Mon Jul 29 11:54:42 2019
             State : clean, degraded
    Active Devices : 4
   Working Devices : 4
    Failed Devices : 0
     Spare Devices : 0

Consistency Policy : resync

              Name : afa_primary:0  (local to host afa_primary)
              UUID : a4ed7a9d:dae6864e:3d03a4ed:6c0d7145
            Events : 17

    Number   Major   Minor   RaidDevice State
       0     259       27        0      active sync   /dev/nvme1n1p1
       1     259       15        1      active sync   /dev/nvme2n1p1
       2     259       33        2      active sync   /dev/nvme3n1p1
       -       0        0        3      removed
       -       0        0        4      removed
       -       0        0        5      removed
       6     259       10        6      active sync   /dev/nvme7n1p1
       -       0        0        7      removed
You have new mail in /var/spool/mail/root
[root@afa_primary ~]#

Check blkid and it lost

[root@afa_primary ~]# blkid  /dev/nvme4n1p1
/dev/nvme4n1p1: PTTYPE="gpt" PARTLABEL="p1" PARTUUID="2149f139-6b8c-4a85-918f-4150fbbb8931"
[root@afa_primary ~]#

X.Niu

Posted 2019-07-29T06:13:18.487

Reputation: 11

Hello and welcome to the forum. This link might help you. It is about the difference between blkid and mdadm. https://unix.stackexchange.com/questions/129497/difference-between-uuid-from-blkid-and-mdadm

– Diblo Dk – 2019-07-29T11:12:24.467

Thanks for you answer. What I want to say is that, when mdadm creating a raid, it produce a UUID and write it into its disk members. So use blkid, I can check that a disk is a member of a raid. When I reboot the system, the UUID in disk is lost. I had search manage websites and has no idea what is the reason and how to solve it or avoid it. – X.Niu – 2019-07-30T03:12:37.633

I'm sure the raid hasn't degraded due to the lack of UUID. It is only a problem to generate the UUID if the partitions cannot be accessed. Therefore, look for problems in communication between blkid and drives. eg. kernel, cables, bad partitions, bad disk, motherboard limits and the like. – Diblo Dk – 2019-08-07T01:51:56.230

No answers