Hi Steve,
as I recognized so far, you haven't patched the kernel yet. Am I right?
If so,
you have to patch the kernel with the patch for the new raidtools. You
will
get it from www.de.kernel.org for example. There is one for kernel
2.0.36 and
2.2.3. Add the patch to the kernel, recompile it and restart your
system.
After the restart you will be able to initialize the raid with
mkraid --really-force /dev/md0
I had the same problem before patching. You can the type
raidstart /dev/md0 and watch the syncing process in /proc/mdstat
Make an mke2fs /dev/md0 and mount it, I think you will be able to use it
immediately while syncing disks.
By the way, the 0.90 are working great.
Dietmar
P.S: I am running kernel 2.2.3 on suse linux 6.0 with only the raidpatch
applied to the kernel
Steve wrote:
> Hiyas All,
> Have been trying for the past few day to setup RAID-1 on a
> redhat linux system and came across some weirdnesses..
>
> firstly, this is the system its running on..
>
> RedHat Linux 5.2
> raidtools 0.90 (the rpm from the redhat contrib site - originally started
> out with the 0.5 version that came with the distribution)
> Two SCSI hard drives, 4Gb each (these are identical drives)
> the hardware is a compaq proliant 800 system (NCR 53c8xx SCSI Bus)
>
> originally, i followed all the docs, and suceeded in mirroring the two
> drives, but the problems started when i went to format the partition
> (/dev/md0) with mke2fs, i got an error along the lines of "the file system
> returned an unsupported library function call" or something, i no longer
> can duplicate this message to give the exact wording because in my efforts
> to fix it it got worse (heh, usually the case when you dont really know
> what yer doing)
>
> I then thought "ah-ha ! mabee i need to update the mke2fs program" so i
> downloaded the rpm for e2fsprogs 1.14 and updated it - still the same
> problem, i then decided that mabee it had something to do with the
> raidtools package, and so went and updated that rpm to the one on the
> contrib site as well (raidtools-0.90-19990309) this produced even more
> confusion due to the fact that many of the nifty little toys i had been
> playing with (ckraid or something) seemed to dissapear, and it STILL didnt
> fix the problems.. so, i downloaded the sources and compiled my own copy of
> the raidtools package, this gained me the readme docs which talked about
> the changeover from the md* commands to the raidstart/stop/other stuff
> commands - yet it STILL didnt work.. "blow it" I thought, and wiped both
> drives clean and prepared to start again, only.. this time, i made one big
> partition and went from there, after rebooting to clear everything out, i
> went back and used the default raid-1 sample config and changed the spare
> drive number to 0 (I didnt have any spare drives) and ended up with the
> following /etc/raidtab file..
>
> raiddev /dev/md0
> raid-level 1
> nr-raid-disks 2
> nr-spare-disks 0
> chunk-size 4
> persistent-superblock 1
> device /dev/sdb1
> raid-disk 0
> device /dev/sdc1
> raid-disk 1
>
> this seemed all well and good, and i fdisked the drive, insmoded the
> raid1.o module and prepared to re-mirror my drives..
>
> it then returned the following error and i cant seem to make any headway
> after this..
>
> [root@xxxxxx raidtools-0.90]# mkraid --force /dev/md0
> DESTROYING the contents of /dev/md0 in 5 seconds, Ctrl-C if unsure!
> handling MD device /dev/md0
> analyzing super-block
> disk 0: /dev/sdb1, 4188937kB, raid superblock at 4188864kB
> disk 1: /dev/sdc1, 4188937kB, raid superblock at 4188864kB
> mkraid: aborted
>
> if i dont include the --force it just gives me the "aborted" message
>
> I am really stuck here and have tried all day to find some documentation on
> the web reguarding this and the mke2fs problem - any help would be mightily
> appreciated..
>
> thanks in advance
>
> Steve.
>
> PS : I know i cant spell.. and i really dont mind re-compiling anything if
> redhat in their infinite wisdom "accidently" left something out of their
> rpms :)
>
> on another note also - this mke2fs problem was also bugging another guy i
> met that was trying to do the same thing on another redhat system, is this
> a common problem ?
