Affinity Modes

In this section, the various modes available for colocation of related data and computation is discussed.

You can run the SnappyData store in the following modes:

  • Local Mode: Used mainly for development, where the client application, the executors, and data store are all running in the same JVM

  • Embedded SnappyData Store Mode: The Spark computations and in-memory data store run colocated in the same JVM

  • SnappyData Smart Connector Mode: Allows you to work with the SnappyData store cluster from any compatible Spark distribution