we have just succeeded in the task of configuring our first GAD volume, with a two G200 and one HUS110.
In our tests, we successfully used one CCI with two instances of HORCM to manage both storage systems via FC.
While in the GAD documentation there are plenty of graphs showing each CCI managing both storage systems, using a couple of command devices (via FC), our partners prefer the topology where each CCI manages just one storage system via FC, the two resort to IP for pair operations control.
We asked them why mix FC and IP for CCI operations and their best answer goes like this: having one CCI with two instances has the disadvantage of raising up the probability of errors, sending commands to the wrong storage system; in the two CCI configuration, you can't send commands to the wrong G200, at least if your are sober enough to login into the right CCI system.
This explanation seems to justify enough the mixing of FC and IP.
In your opinion, which configuration is best, and why?