Hi all,
I am ... let's say a sys admin, and I have this little problem every time I build a mirror with mdadm. I build it by the ubuntu install and there come the problems. The ubuntu version is 8.04.1, but it happens with 7.10 too. I am a little scared about the data on this raid because i use it everywhere. So this is the exact problem: install, configure raid1 with sda,sdb and the folowing message appears on random basis at the start up screen (this is from the syslog file but the same is shown at the start).

md: md0 stopped.
Aug 25 09:52:54 SupportivoFileserver kernel: [ 36.004675] md: invalid raid superblock magic on sda
Aug 25 09:52:54 SupportivoFileserver kernel: [ 36.004722] md: sda does not have a valid v0.90 superblock, not importing!
Aug 25 09:52:54 SupportivoFileserver kernel: [ 36.004728] md: md_import_device returned -22
Aug 25 09:52:54 SupportivoFileserver kernel: [ 36.004960] md: bind<sda2>
Aug 25 09:52:54 SupportivoFileserver kernel: [ 36.078227] md: md0 stopped.

After a huge research I didn't find any solution, except that the metadata has to be set to v0.90 so the kernel could work properly. I checked the version and it IS 0.90. I put it also in the mdadm.conf file as a declaration but with no result. I switched the motherboard, assuming it could be the chipset/driver issue, I switched the PSU but with no result. Tested all the HDDs with Seatools and no problems (also they are almost new). I have no ideas what is happening. I would be very grateful if you help me with this!
Thanks!

If you run into issues with using whole drives, label your drives, and for each drive, create a partition and put all the space into that partition.

Then mirror the partitions instead.

If I understand you right, this is what the ubuntu server installer does: partition as FD, mirror them, then make file system on top of md*. I think there is no difference between ubuntu installer and manual mdadm create... cfdisk... and mkfs.reiserfs... commands. But after all this is the strange thing, that the manual built raids are out of the error logs. If you need more info please ask!
Edit: I use raid only on a "data" partition, not the whole hdd. The system is outside the raid.
root@Fileserver:~# fdisk -l /dev/sda

Disk /dev/sda: 160.0 GB, 160041885696 bytes
255 heads, 63 sectors/track, 19457 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Disk identifier: 0x02820282

Device Boot Start End Blocks Id System
/dev/sda1 * 1 1094 8787523+ 83 Linux
/dev/sda2 1095 19330 146480670 fd Linux raid autodetect
/dev/sda3 19331 19457 1020127+ 82 Linux swap / Solaris

root@Fileserver:~# fdisk -l /dev/sdb

Disk /dev/sdb: 160.0 GB, 160041885696 bytes
255 heads, 63 sectors/track, 19457 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Disk identifier: 0x29cd29cc

Device Boot Start End Blocks Id System
/dev/sdb1 * 1 1094 8787523+ 83 Linux
/dev/sdb2 1095 19330 146480670 fd Linux raid autodetect
/dev/sdb3 19331 19457 1020127+ 82 Linux swap / Solaris

Hmm, my bad, didn't see the reference to sda2. Problem you have usually happens to whole drives, first I see this with mirrored partitions.

Try zeroing out the last 128K of sda2, where the superblock is, and try again. See:

http://www.linuxquestions.org/questi...-drive-661308/

And zero out last 128k of sda2, not sda.

Before you do this, make sure sdb has a copy of all the data on sda, or better yet, all your data is backed up.