How to implement controller fault tolerance for a pool

We are evaluating/testing our new ME-4024. We set up a 24 disk system with A & B pools using ADAPT (12 disks per pool). It appears from testing that each of the A/B pools is served by only one of the controllers. For instance, when I reset the ports on controller A [A0, A1], service is interrupted for all the consumers of volumes in pool A. I conclude that controller B is not serving pool A in any backup capacity. 

Is there a system configuration that implements an active/backup controller like service for the pools, so that if one controller fails in any way, the other controller will serve all pools?