Our terminology has changed a few times over the years, so the confusion is understandable.
I will start with the Data ONTAP 7.2.x and 7.3.x families. Here we use the term active/active configuration for a single system with two storage controllers. Each controller normally serves its own data. If one controller fails, the partner controller takes over and serves both its own data and the data for the partner. Physically, the two controllers are connected to each other. Each controller also has one or more connections to its own disks and to its partner's disks.
In Data ONTAP 7.1 and earlier releases, this same configuration above was called a cluster.
Unlike some storage vendors, NetApp does not use passive controllers.
Starting with Data ONTAP 8.0 Cluster-Mode, we again use the term cluster. Here is the official definition:
In Data ONTAP 8.x Cluster-Mode, a group of connected nodes (storage systems) that share a global namespace and that you can manage as a single virtual server or multiple virtual servers, providing performance, reliability, and scalability benefits.
Note that we are no longer restricted to a pair of storage controllers. You can scale the Data ONTAP 8 cluster to include more controllers as needed. However all of the controllers are still active and available to serve data.
If you order a storage system with two controllers (there are applications where a single-controller configuration is adequate), the "default" would be that you have an active/active configuration or cluster. However, you need to be sure everything is configured correctly so that the system can fail over between controllers successfully.