Hi, all. I have a cluster of 2 servers. We configured a couple of extra
interfaces on them and rebooted the machines, now it's claiming the
metadb for my disk group is gone. The cluster was talking over
interfaces ce4 & ce8, we configured interfaces ce3 & ce7, then
performed an scshutdown and rebooted the server. Now, during the boot,
I'm getting the errors:


Oct 10 12:07:04 utsocs1 Cluster.Framework: stderr: metaset: utsocs1:
dg-smp: there are no existing databases
Oct 10 12:07:04 utsocs1 Cluster.Framework: stderr: metaset: utsocs1:
dg-smp: must be owner of the set for this command
mount: I/O error
mount: cannot mount /dev/md/dg-smp/dsk/d120

The configuration of those 2 interfaces should have nothing to do with
the metadb, so what could have happened to it?

Previously, when I looked at the disks via format I saw the the quorum
disk was reserved:

       0. c0t8d0 <drive not available: reserved>
          /pci@1c,600000/scsi@1/sd@8,0


but that is no longer the case:
       0. c0t8d0 <SUN72G cyl 14087 alt 2 hd 24 sec 424>
          /pci@1c,600000/scsi@1/sd@8,0

Any time I try to perform an action on the resource-group I get the
following message:
scswitch: rg-smp: resource group is undergoing a reconfiguration,
please try again later

And the scstat -g shows:
Group: rg-smp              utsocs2             Pending online


Is there any way to revive my metadb and disk group. I don't know what
happened, but I have an application running on that partition and I
need to know how to get it back. Any help would be appreciated.

Thanks...

In article <1160500408.102756.14110@k70g2000cwa.googlegroups.com>, bozothedeathmachine <bozothedeathmachine@gmail.com> wrote:
> Hi, all. I have a cluster of 2 servers. We configured a couple of extra
> interfaces on them and rebooted the machines, now it's claiming the
> metadb for my disk group is gone. The cluster was talking over
> interfaces ce4 & ce8, we configured interfaces ce3 & ce7, then
> performed an scshutdown and rebooted the server. Now, during the boot,
> I'm getting the errors:
>
>
> Oct 10 12:07:04 utsocs1 Cluster.Framework: stderr: metaset: utsocs1:
> dg-smp: there are no existing databases
> Oct 10 12:07:04 utsocs1 Cluster.Framework: stderr: metaset: utsocs1:
> dg-smp: must be owner of the set for this command

Check into metaset stuff. 'man metaset' for more information; it also
talks about cluster handling of SVM metasets as well as how to deal with
recovery in this kind of situation, I think.

See the -j or -t option in the man page for metaset and read carefully.
(And also read the whole man page, and not just these options alone.)

Also ensure that the appropriate SVM daemons (req'd by metaset) are
running and can communicate between the hosts, too.

-Dan

> Check into metaset stuff. 'man metaset' for more information; it also
> talks about cluster handling of SVM metasets as well as how to deal with
> recovery in this kind of situation, I think.
>
> See the -j or -t option in the man page for metaset and read carefully.
> (And also read the whole man page, and not just these options alone.)
>

Thanks for the reply. I have tried to take ownership of the disk group,
but I get some error about a DCS (whatever that is).

utsocs2 root#metaset -s dg-smp -t
metaset: failed to notify DCS of take

I get this message both with and without the force option.

> Also ensure that the appropriate SVM daemons (req'd by metaset) are
> running and can communicate between the hosts, too.

Those are the "meta" commands configured in the /etc/inetd.conf
directory, correct? If so, they are active.

# rpc.metad
100229/1        tli     rpc/tcp         wait    root
/usr/sbin/rpc.metad     rpc.metad
# rpc.metamhd
100230/1        tli     rpc/tcp         wait    root
/usr/sbin/rpc.metamhd   rpc.metamhd
# rpc.metamedd - DiskSuite mediator
100242/1        tli     rpc/tcp         wait    root
/usr/sbin/rpc.metamedd  rpc.metamedd

Thanks again,
Ben..
# rpc.metacld - DiskSuite cluster control
100281/1        tli     rpc/tcp         wait    root
/usr/sbin/rpc.metacld   rpc.metacld