[SunHELP] Problem with SunCluster 3

Hi Cluster gurus,
I have a problem with a two nodes suncluster 30...

> The two nodes of cluster have nafo conected to diferent switches (switch
> A, switchB), and those switches are connected separately to other two
> switches (switchAA, switchBB). So, switchA is connected to switchAA and
> switchB to switchBB. Between switch A and switch B there is not physical
> link but switchAA and switchBB are linked for redundancy porpouses.
> 
> The problem is that if the cable among switchA and switchAA is broken, the
> cluster node would not detect that nafo can not comunicate with services
> and unless they make a ping to a non cluster ip (ie and Ip out of the
> subnet) to realize that this link is not able to go outside the nafo will
> not be migrated to the backupinteface and the services will not be
> available from outside.
> 
> 
> Is any posibility of forcing a ping to the defaulrouter ip_addres during
> the 'it's alive' internal procedures of Suncluster? This would be a
> solution (we have seen that the cluster sw make pings to diferent
> ip_addresses in order to check the quality of the nafos).
> 
> I hope you understood the scenario.
> Best regards