Where is my data actually stored?

Treating the data as objects allows MatrixStore to hide the management of the location of data within the cluster. MatrixStore automatically finds the most appropriate location in the cluster to store an object. This dramatically reduces the management overhead of storing data or adding capacity to a large archive.

Applications treat MatrixStore like a black box;

They send data to it with no knowledge of where the data eventually resides. Upon completion of a successful write operation the application receives a globally unique Identifier with which it can retrieve the data at any time.

MatrixStore does NOT address the data by its contents. (google for more information on CAS weakness)

The location of the object within the cluster may change due to internal load balancing or data protection management yet the globally unique key remains the same.