Question about primary data source for distributed environment

My understanding ...

Multi-Gateway: multiple machines each with its own full installation of Appian & all the engines. This is a high availability solution. In theory, if one machine goes down, the other(s) will take over and still be running all the engines.

Distributed Gateway: multiple machines each with its own installation of Appian with a subset of the engines so that together they run all the engines. This is a load balancing solution. This way if one machine is not able to handle the full load of all 16 engines, a subset of the engines may be split off to another server.

In a Multi-Gateway, the servers all access the same data source. I would think the same would have to be true in a distributed gateway architecture; but, I'm looking for confirmation.

TIA,
tdw

My understanding ...

Multi-Gateway: multiple machines each with its own full installation of Appian & all the engines. This is a high availability solution. In theory, if one machine goes down, the other(s) will take over and still be running all the engines.

Distributed Gateway: multiple machines each with its own installation of Appian with a subset of the engines so that together they run all the engines. This is a load balancing solution. This way if one machine is not able to handle the full load of all 16 engines, a subset of the engines may be split off to another server.

In a Multi-Gateway, the servers all access the same data source. I would think the same would have to be true in a distributed gateway architecture; but, I'm looking for confirmation.

TIA,
tdw