Richard and Mark,
Thank you for your assistance. We have verified all of your
recommendations and re-defined the volume group per pvcreate, vgcreate,
and lvcreate. Once complete, the volume group is recognized and I is
used, in our case for Oracle user data. Our linux operating system and
Oracle binaries are on 3390 DASD and have no issues. We are still
experiencing the same issues when we re-IPL the linux guest. We don't
think that it is a hardware problem, since this is the same disk string
used for our SUSE 10 guests. Can you think of anything else that we
might have missed, particularily if it is unique to SUSE 9 (as opposed
to SUSE 10)
Thank you, Bob
-----Original Message-----
From: The IBM z/VM Operating System [mailto:[email protected]] On
Behalf Of Richard Troth
Sent: Friday, March 06, 2009 7:15 PM
To: [email protected]
Subject: Re: fcp disk under SUSE9
We use FCP with SLES 9 all the time.
Is your disk on two paths? If so, then you also need to have MPIO
installed and running. (multipathd and boot.multipath)
You ran mkinitrd and zipl, but did you notice if they were successful?
Is /boot on its own disk? Is it mounted RW? Is the underlying minidisk
RW?
You said that you have to define the vol group manually. That's how we
always do it: pvcreate, vgcreate (or vgextend), and lvcreate (or
lvextend).
I recommend that you NOT partition a SAN volume intended to be used as a
PV. (ie: pvcreate /dev/sda instead of pvcreate /dev/sda1) But I don't
think that is your problem. More inclined to think there is a hardware
problem, but how then does SLES 10 cope?
I hope this helps.
On 3/6/09, Robert J McCarthy <[email protected]> wrote:
> We are attempting to use zfcp scsi devices with SuSE 9 sp4 (linux on
> z) We place a 10gb lun into a volume group.
> We have to manually create the volume groups since yast is not working
> for SUSE 9 (known issue) Once this procedue is completed, the disk is
> recognized and can be used for read/write operation. The mount is
> placed in fstab.
> After we reboot the machine nothing comes back (We perform mkintrd and
> zipl).
> We can manually re-define the volume group and the system will
> recognize it again.
> There are occasions when the data is corrupted. Has anyone run into
> issues attaching fcp disk to a volume group with SUSE 9 ? We have had
> no problems utilizing fcp disk with SUSE 10.
>
--
Sent from Gmail for mobile | mobile.google.com
-- R; <><
Q: Isn't filtering stupidity elitist?
A: Yes. Yes, it is. That's sort of the whole point. -- Ortiz and Starr
