The fuel-slave-1 node running cinder services is unable to ping my backend storage array even though the fuel-master is able to reach it correctly.

I have cinder services running on a slave node. On this slave node, cinder.conf points to my storage array as the backend. However, the slave node is unable to reach my array - the master is able to reach the array without any issues. I thought that the slave only spoke to the master which did all the resolving for it. Should I be doing any IP forwarding on the slave node as I had to do for the master?