Hi there,

having a HACMP ES 4.4.1 running some day an operator did switch off both
the machines. Since then I have some errors I cannot recover. First of
all there is a service adapter defined. Neither smitty show the adapter
under 'Delete an adapter' oder 'Cluster verification' nor is the address
configured when taking the ressource group online. But if I try to
define it, smitty says the adapter already exists.

The other strange behavior is one of the volume groups. If verified with
smitty cluster verification the output tells:

Performing volume group consistency check.
------------------------------------------
ERROR: Shared VG BMC_vg not found on node ms3hv02

But later in the log, all seems to be found:

Verifying Resource Group: netmanag2_rg
-------------------------

Verifying Resources on each Node...


Verifying SERVICE_LABEL: ha3hv01_bmc_svc
Verifying Filesystem: /ServiceCenter
Verifying Filesystem: /controlm
Verifying Volume Group: SC_vg
Verifying Volume Group: BMC_vg

The VG is mounted and moved whith the ressource group as one would expect.

Has anyone an idea how to overcome this without re-defining the whole
cluster? Would applying an old snapshot solve the problem?

Any help is highly appreciated,
Patrick

Patrick <rhodan_nospam_5625@freenet.de> wrote in message news:<bk1ljg$gvi$1@online.de>...
> Hi there,
>
> having a HACMP ES 4.4.1 running some day an operator did switch off both
> the machines. Since then I have some errors I cannot recover. First of
> all there is a service adapter defined. Neither smitty show the adapter
> under 'Delete an adapter' oder 'Cluster verification' nor is the address
> configured when taking the ressource group online. But if I try to
> define it, smitty says the adapter already exists.
>
> The other strange behavior is one of the volume groups. If verified with
> smitty cluster verification the output tells:
>
> Performing volume group consistency check.
> ------------------------------------------
> ERROR: Shared VG BMC_vg not found on node ms3hv02
>
> But later in the log, all seems to be found:
>
> Verifying Resource Group: netmanag2_rg
> -------------------------
>
> Verifying Resources on each Node...
>
>
> Verifying SERVICE_LABEL: ha3hv01_bmc_svc
> Verifying Filesystem: /ServiceCenter
> Verifying Filesystem: /controlm
> Verifying Volume Group: SC_vg
> Verifying Volume Group: BMC_vg
>
> The VG is mounted and moved whith the ressource group as one would expect.
>
> Has anyone an idea how to overcome this without re-defining the whole
> cluster? Would applying an old snapshot solve the problem?
>
> Any help is highly appreciated,
> Patrick

Since you know how the cluster should look, redefining it should take
less than an hour; almost certainly quicker that messing around.
Applying an old snapshot should also work (as long as it isn't too old
of course).

Iain.