Cloud Objects: Programming the Cloud with Object-Oriented Map/Reduce

Cloud Objects parallelizes Object-Oriented programs written in Java using Map/Reduce by adding simple, declarative annotations to the code. The system automatically persists objects to a partitioned filesystem and efficiently executes methods across the partitioned object data. Using Cloud Objects, data-intensive programs can be written in a simple, readable, object-oriented manner, while maintaining the performance and scalability of the Map/Reduce platform. Cloud Objects shows that it is possible to combine the benefits of an object-oriented model and the power of Map/Reduce. Keywords-Map/Reduce, Hadoop, JPA, Cloud Computing