Error creating Raid 5

[Jocko]

So nothing on the SMART side.

However I think that you have an issue with two disks (sdc and sde).

Did you try this?

Can you try to do a raid5 with 3 devices with one missing: missing(sda8)-sdb8 and sde8 and see what happens when you format it

[b1gnas]

Did you try this?

Can you try to do a raid5 with 3 devices with one missing: missing(sda8)-sdb8 and sde8 and see what happens when you format it

Not yet.

Is it correct to have 4 disks with gpt partition and the first with MBR? Or is better to format again in MBR

Anyway, if the raid with 3 disks fails, is it possible to restore the original firmware using the fvdw console?

[Jocko]

Hi b1gnas,

anyhow you will be able to restore the Lacie firmware.

[b1gnas]

Ok, where can i find a guide to restore the original firmware

Now i'm trying to create a raid with 3 disk...but if it fails i think i will restore the original

[b1gnas]

when i create partitions on the disks i have this result

Device Boot Start End Blocks Id System
/dev/sdb4 201 121600 975145500 5 Extended
/dev/sdb5 209 312 835380 83 Linux
/dev/sdb6 321 328 64260 83 Linux
/dev/sdb7 337 400 514080 83 Linux
/dev/sdb8 409 121600 973474740 83 Linux

and not

Device Boot Start End Blocks Id System
/dev/sdb4 201 121600 9xxxx+ 5 Extended
/dev/sdb5 209 312 803218+ 83 Linux
/dev/sdb6 321 328 8001 83 Linux
/dev/sdb7 337 400 514048+ 83 Linux
/dev/sdb8 409 121600 9xxxx+ 83 Linux

like in the first install instruction for the 5big1, is it a problem?

[fvdw]

Your output is correct, one blok is if I remember well 8 MB. So 8 blocks are 64 MB.

[b1gnas]

same error with 3 disks :sob

Code: Select all

root@fvdwsl-base:/ # mdadm --create /dev/md0 --level=5 --raid-devices=3 missing
/dev/sdb8 /dev/sde8
mdadm: /dev/sdb8 appears to contain an ext2fs file system
    size=973474740K  mtime=Mon Apr 20 22:07:41 2015
mdadm: /dev/sde8 appears to contain an ext2fs file system
    size=973474740K  mtime=Mon Apr 20 22:07:41 2015
Continue creating array? y
mdadm: array /dev/md0 started.
root@fvdwsl-base:/ # cat /proc/mdstat
Personalities : [raid0] [raid1] [raid6] [raid5] [raid4] [raid10]
md0 : active raid5 sde8[2] sdb8[1]
      1946949248 blocks level 5, 64k chunk, algorithm 2 [3/2] [_UU]
     
unused devices: <none>
root@fvdwsl-base:/ # cat /etc/mdadm.0 > /etc/mdadm.conf
root@fvdwsl-base:/ # mdadm --detail --scan >> /etc/mdadm.conf
root@fvdwsl-base:/ # mke2fs -t ext3 -j -m 1 /dev/md0
mke2fs 1.42.11 (09-Jul-2014)
Creating filesystem with 486737312 4k blocks and 121692160 inodes
Filesystem UUID: 2d23765d-2b22-431a-a2dd-b4cf98c82942
Superblock backups stored on blocks:
   32768, 98304, 163840, 229376, 294912, 819200, 884736, 1605632, 2654208,
   4096000, 7962624, 11239424, 20480000, 23887872, 71663616, 78675968,
   102400000, 214990848

Allocating group tables: done                           
Writing inode tables: done                           
ext2fs_mkdir: Attempt to read block from filesystem resulted in short read while creating root dir


[fvdw]

I did not read the full thread, but did you try this command to see if we have a faulty disk ?

Code: Select all

mdadm --detail /dev/md0

[Jocko]

There is a strange thing if the mdstat file: no metadata is reported :scratch

The 2d line should be

Code: Select all

1946949248 blocks super 1.0 level 5,....

and not

Code: Select all

1946949248 blocks level 5,.....

[Jocko]

Hi b1gnas,

Fvdw has compiled a new kernel version for 5big1. So to check if it is a kernel issue, you can try it.

Download the 2 archives in the share fvdw (I assume that you already created it). Unzip the two files

So In Putty do

Code: Select all

mkdir /lib/modules/3.14.2
(check if it is here )
ls -l /lib/modules

Untar the modules in the folder

Code: Select all

tar -xf /share/1000/fvdw/5big1-3142-1.tar -C /lib/modules/3.14.2

Now, load the new kernel version

Code: Select all

dd if=/share/1000/fvdw/lib/modules/UIMAGE-3142-5big1-1 of=/dev/sda6

Destroy your raid and reboot.

I hope that this time you will succeed to create the RAID.

Please to post the dmesg output.