Good Night All,
I am presently having this issue with a Checkpoint cluster operating in load sharing multicast mode connected to a nexus 5500 series cluster core.
When in this mode all switches including the nexus are not able to arp to the VIP of the respective firewall interfaces. So what is occurring is that the DG of the nexus as well as 3com switches in the customer's environment is having an incomplete ARP entry for the Firewall VIP.
I have seen in stated that RFC 1812 does not allow this. Can you advise if this is the reason why all switches with a layer 3 interface are not able to arp to the respective DG on the Firewall cluster? But hosts have no issues since this will make a compelling case for my customer to remain with the checkpoint unicast mode which is working with no issues?
Are there any workaround for multicast mode beside entering manually arp entries on the switches to point to the multicast mac?
We are trying to explain to the customer that this is the way forward however they are of the belief that the core switch (nexus) is the issue and once resolved there will allow the other switches 3com to communicate without the need of the multicast mac static entry.
Please help!!! I have seen much discussions here and we presently have a cisco tac case opened but no real root cause determined by cisco
