Re: Anyone got raid to work on a sparc?
On Wed, Sep 15, 1999 at 07:44:57PM +0200, Anders wrote:
> I'm trying to set up raid0 with 2 disks.
> In short what i've done is creating /etc/mdtab with
> mdcreate raid0 /dev/md0 /dev/sdb1 /dev/sdc1
> this gives a /etc/mdtab containing
> /dev/md0 raid0,4k,0,f9b7b5b6 /dev/sdb1 /dev/sdc1
> >From what i've read in howto's and man pages i should now be able to just
> type mdadd -a to add /dev/md0 and i've also tried 'mdadd /dev/md0
> /dev/sdb1 /dev/sdb2'
> both of these gives the error,
> /dev/sdb1: Invalid argument
> /dev/sdc1: Invalid argument
> I've tried raidtools 0.42 and 0.41(0.41 is the only one i'we been able to
> The kernel version i'm using is 2.2.12
> Computer is an ELC
> Would I be better off with another kernel version or another version of
> Anyone else had problems with this? Any suggestions for how to get it to
I got it to work (raid 0, raid 5) with the new raid patches and
raidutils (Get potato source package and build for slink)
I had a couple of problems with crashes, oopses and the like so i would
currently not recommend relying on working raid on sparc ...
Also i used the "qlogicpti" SBUS Wide SCSI Controller which seems
to have some problem with many Drives on the same bus or the kind.
I happens to me that i was not able to resync a raid5 with 6 drives
in an external Sun Multipack.
The same Multipack (Same cables etc) workes flawlessly on my i386
Linux box with an NCR based PCI Wide controller. (Same raid
patches + raidtools + same kernel version)
Florian Lohoff email@example.com +49-5241-470566
... The failure can be random; however, when it does occur, it is
catastrophic and is repeatable ... Cisco Field Notice