Redundant sonicwalls, redundant switches underneath

[Kooch]

I have two Sonicwall NSA 4500s linked together for high availability. Underneath these I have two Procurve 2810 switches which are stacked. Is there any way to setup the Sonicwalls so that in the event of one of the HP switches dying there is little to no disruption of traffic?

As it stands right now the primary 4500 is connected to the top Procurve and the secondary 4500 is connected to the bottom. In the event of a Sonicwall dying the other will take over, but if the top switch dies traffic will grind to a halt.

I was thinking of setting up interface monitoring on the LAN interface pinging something downstream, so if the top switch dies the secondary Sonicwall takes over. However, my boss thinks there is a better way to do this using bridging mode on another interface. That doesn't seem to be what the bridging feature was designed for. Is this a solution worth investigating even?

Thanks in advance.

 

[sstoppel]

It is the "stacked" part that is the gotcha. Can you just get a cheap unmanaged switch and put the failover in that?

 

[Kooch]

Do you mean add a cheap unmanaged switch somewhere downstream and use it as the target for the interface monitoring? Or do you mean to use it in a different manner?