Overview (DeployR 7.3.0 Java RBroker Framework)

Packages
Package	Description
com.revo.deployr.client.broker	Defines a high-level programming model for DeployR-enabled client applications.
com.revo.deployr.client.broker.app	Defines interfaces for `RBroker` compliant client application simulators.
com.revo.deployr.client.broker.config	Defines `RBroker` configuration options.
com.revo.deployr.client.broker.options	Defines `RTask` configuration options.
com.revo.deployr.client.factory	Defines convenience factory classes for DeployR-enabled client applications.

Java RBroker Framework ~ Simple R Integration for Java Application Developers

The Java RBroker Framework exposes a high-level programming model for building DeployR-enabled client applications. By using the RBroker Framework an application developer can focus entirely on integrating R Analytics, while offloading the complexity of managing client-side API task queues and server-side R session lifecycles.

The basic programming model for working with the framework is as follows:

Decide if the R Analytics tasks for your application should execute as:
1. Discrete tasks: authentication optional, grid resources allocated at runtime, results returned immediately, no persistence. Good for prototyping and public facing production deployments.
2. Pooled tasks: authentication required, grid resources pre-allocated, results returned immediately, optional persistence to repository. Good for enterprise production deployments, consistent runtime, high-throughput environments.
3. Background tasks: authentication required, grid resources allocated at runtime, results persisted for later retrieval. Good for periodic, scheduled or batch processing.
Use the RBrokerFactory to create an appropriate instance of RBroker.

Use the RTaskFactory to create one or more RTask for your application.

Submit your RTask to RBroker for execution.

Track the progress of your RTask using RTaskToken.

Integrate the results of your RTask found within RTaskResult.

This programming model can be further simplified for application developers by leveraging asynchronous callbacks. Simply register an RTaskListener with your RBroker instance and the RBroker will automatically notify your application when each RTask completes. This approach allows your application to skip step 5. above and scale effortlessly.