1500D Cluster (a-p) HA Failover Performance w/Seesion Pickup Enabled

[Camshaft007]

All So I' m in the process of trying to squeeze out the best bang for my buck with our new 1500D Cluster(s) and I would like to have our Failover performance setup as optimal as I can. Do you think using two copper 1GB HA ports with session off loading to a single 10GB port would be optimal? (FYI as stated in the subject; Session Pickup will be Enabled) Anyone aware of any gotcha' s with the Gratuitous ARPs going to a Nexus7K that might cause a delay in failover scenarios? One caveat to this whole shebang is that all my Policies have UTM profiles assigned on them, and from what I understand, UTM sessions will get dropped and need to be restarted, is that correct? I' d prefer not to mess with " Age" and " Override" as much as possible and just keep it as close to default as possible. Any advice would be seriously appreciated!