actually it can be done through the Traps if the device are configure and supporting to send such of failover traps to NNM
for the 2 FW ip address having Same can be solved if you changed the MGMt ip address on the devices to have such as loopback or virtual interfaces to manage the devices through and don't forget to exclude the Shared physical interface in the discovery
what you can do as the scenario below
Physical IP address 10.0.0.1
to test the failover functionality make the first down "FW1 Down" , in the mean time you can do the following to discover all the traps generated on the FW2 from the below command .
nnmtrapdump.ovpl -source FW2 >/tmp/trap.txt
Hence FW2 is the Hostname or the device name in NNM as configured .
Browse the trap.txt file for all the traps generated by Fw2 , and check if there are any trap for Failover activity like FW2 active or primary
it's easy to be checked from the trap varbind .
Note : kindly make sure that all the supported Mibs and Traps are loaded in NNM .
let me know if it's working with you or not .
All the best . Regards Mostafa Hassan HP AIS NNM-NA-OO CCNA-CCNP-ITIL-VCA-Cloud-VCA DataCenter
There has to be a unique criteria by which each portion of the redundant configuration can be communicated with... An IP address thats the same is not unique so when NNMi polls the shared IP address , the chassis inventory wont match both portions of the HA cluster, just which ever was discoverd.