I see you opened a Support the next day, which guarantees proper followup
on you issue.

For others monitoring this group, issues such as this typically require more
in-depth investigation. It helps to get a vxexplore from each node so we
may more closely review the configuration.

For expedience, you may retreive the utility here:

ftp://ftp.veritas.com/pub/support/vxexplore.tar.Z

extract, run, and return results to ftp.veritas.com:/incoming . Use the
case number when prompted, and answer "no" to restarting vxconfigd in debug
mode.

Best Regards,

"Tomasz Kilinski" wrote:
>Hello,
>I have a problem with cvm master failover.
>Configuration:
>hw: Sun Netra200, HBA Qlogic2200, Brocade 3800, HDS 9200.
>sw: Solaris8, SANPointFS HA 3.4, VCS2.0 + patch04 - 4 node cluster.
>I have configured ClusterFileSystem with shared disk group. Everything
>looks/works fine but ...
>If I disconnect node (HBA/cable/switch port failure) which is primary for
>CFS I have all cvm service groups failed on all nodes. I expect that another
>nodes should became primary.
>If I disconnect node which is NOT primary node for CFS other nodes works
>fine.
>
>Anybody have experience with simillar problem?
>Any suggestions pls.
>
>
>thx and rgds
>
>Tomek
>
>
>