Thanks for reading this.

As many of the BLU members know, I'm the editor of The Telecom Digest, which runs on a donated machine at a major university in Cambridge. Recently, the machine dropped a RAID disk, and is running in "degraded" status. Our benefactors have been kind, and have assigned us a new machine for the Digest: I am installing software RAID1 on it in preparation for adding LVM and copying the Telecom Digest archives over to it.

However - I have to do this by ssh. I've never installed any kind of RAID on a "live" machine, so I'm going /very/ carefully. If I brick the box, I'll have to ask for a reinstall.

As it stands now, I have the machine booted into an ext3 partition that will be used only for the RAID and LVM installs. Once I have the LVM up and running, I'm going to switch to booting from a separate /boot partition (I don't care if /boot is on RAID at this point: the clock is ticking on the "old" box, so I need a RAID1 array for the root first and foremost).

The machine came with a default Ubuntu 13.04 LTS install, which includes an LVM on /dev/sda, and a blank /dev/sdb. The plan is to create a degraded RAID1 array on the "spare" drive, and then copy the "live" drive data into it and subsequently join the two drives together in a RAID1 array, with LVM on top of RAID1.

However, we know what they say about the best laid plans ...

I've managed to create some sort of array, named "md127", but I haven't been able to figure out how. I did an "mdadm --create" with a name of "md1", but I've wound up with "md127". It appears to be working, albeit in degraded mode, but I want to go slowly and figure out what happened before I wind up with a non-standard install which might cause problems later.

Here's the output of my most recent ssh session (separators added for clarity):

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
telecomdigest5@telecom-new:~$ sudo mdadm -D /dev/md1
mdadm: cannot open /dev/md1: No such file or directory
telecomdigest5@telecom-new:~$ sudo mdadm -D /dev/md127
/dev/md127:
        Version : 1.2
  Creation Time : Wed Jun 11 00:23:59 2014
     Raid Level : raid1
     Array Size : 716464832 (683.27 GiB 733.66 GB)
  Used Dev Size : 716464832 (683.27 GiB 733.66 GB)
   Raid Devices : 2
  Total Devices : 1
    Persistence : Superblock is persistent

    Update Time : Wed Jun 11 23:35:19 2014
          State : clean, degraded
 Active Devices : 1
Working Devices : 1
 Failed Devices : 0
  Spare Devices : 0

           Name : telecom-new:md1  (local to host telecom-new)
           UUID : c4e39dd8:541c124b:bd60a9b1:339d6e3e
         Events : 12

    Number   Major   Minor   RaidDevice State
       0       0        0        0      removed
       1       8       18        1      active sync   /dev/sdb2

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
telecomdigest5@telecom-new:~$ sudo mdadm --examine /dev/sdb2
/dev/sdb2:
          Magic : a92b4efc
        Version : 1.2
    Feature Map : 0x0
     Array UUID : c4e39dd8:541c124b:bd60a9b1:339d6e3e
           Name : telecom-new:md1  (local to host telecom-new)
  Creation Time : Wed Jun 11 00:23:59 2014
     Raid Level : raid1
   Raid Devices : 2

 Avail Dev Size : 1432930035 (683.27 GiB 733.66 GB)
     Array Size : 716464832 (683.27 GiB 733.66 GB)
  Used Dev Size : 1432929664 (683.27 GiB 733.66 GB)
    Data Offset : 262144 sectors
   Super Offset : 8 sectors
          State : clean
    Device UUID : 45c648e7:60ae1ec9:c0623dc0:7b06a215

    Update Time : Wed Jun 11 23:35:19 2014
       Checksum : 6286503e - correct
         Events : 12


   Device Role : Active device 1
   Array State : .A ('A' == active, '.' == missing)

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
telecomdigest5@telecom-new:~$ sudo mkfs.ext4 /dev/md1
mke2fs 1.42.9 (4-Feb-2014)
Could not stat /dev/md1 --- No such file or directory

The device apparently does not exist; did you specify it correctly?

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
telecomdigest5@telecom-new:~$ sudo mkfs.ext4 /dev/md127
mke2fs 1.42.9 (4-Feb-2014)
Filesystem label=
OS type: Linux
Block size=4096 (log=2)
Fragment size=4096 (log=2)
Stride=0 blocks, Stripe width=0 blocks
44785664 inodes, 179116208 blocks
8955810 blocks (5.00%) reserved for the super user
First data block=0
Maximum filesystem blocks=4294967296
5467 block groups
32768 blocks per group, 32768 fragments per group
8192 inodes per group
Superblock backups stored on blocks:
32768, 98304, 163840, 229376, 294912, 819200, 884736, 1605632, 2654208,
        4096000, 7962624, 11239424, 20480000, 23887872, 71663616, 78675968,
        102400000

Allocating group tables: done
Writing inode tables: done
Creating journal (32768 blocks): done
Writing superblocks and filesystem accounting information: done

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
telecomdigest5@telecom-new:~$ sudo mkdir /mnt/raid1

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
telecomdigest5@telecom-new:~$ sudo mount -text4 /dev/md127 /mnt/raid1

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
telecomdigest5@telecom-new:/mnt/raid1$ sudo pico /mnt/raid1/test.txt

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
telecomdigest5@telecom-new:/mnt/raid1$ cat /mnt/raid1/test.txt
This is a test file, to check write permissions on the new RAID1 array.

(End of console output)

So, I am able to format the "md127" array, mount it, and write to it: of course, that's not the final state I'm aiming for, but for now I know that it is working with the "md127" name. I don't know how I wound up with the "md127" name instead of "md1", but if there's no danger sign in that name, I'm happy to go to the next step and make "md127" part of the LVM and proceed to add /dev/sda to the RAID1 array.

All advice, warnings, cautions, paranoia, and gotchas welcome. *PLEASE* keep in mind that I don't have access to the machine! Whatever I do, I have to be able to do via ssh: I'm able to reboot it, of course, but every time I do, I'm taking the chance of losing everything on it.

TIA.

Bill

--
Bill Horne
William Warren Consulting
339-364-8487

_______________________________________________
Discuss mailing list
[email protected]
http://lists.blu.org/mailman/listinfo/discuss

Reply via email to