AFF shelf redundancy local syncmirror or Metro Cluster
I'm trying to build a system in a datacenter that customer requires benefits of syncmirror with a new AFF system. As far as I understand there are two ways to do it.
Local syncmirror which controllers in a same box, controllers communicate NVRAM information via internal and fastest speed and use ADP and HA pair can be member of a mixed cluster and use data migration capabilities. I could not find any document that ADP and syncmirror works together on AFF system.
The other choice is Metro Cluster and use FCVI card to communicate which I believe will limit the NVRAM performance and have to use dedicated root aggr that mean more disks and there is no way to member of a mixed cluster right now.