Developing Elastic Software for the Cloud


Developing standalone applications running on a single computer is very different from developing scalable applications running on the cloud, such as data analytics applications that process terabytes of data, Web applications that receive thousands of requests per second, or distributed computing applications where components run simultaneously across many computers. Cloud computing service providers help facilitate the development of these complex applications through their cloud programming frameworks. A cloud programming framework is a software platform to develop applications in the cloud that takes care of nonfunctional concerns, such as scalability, elasticity, fault tolerance, and load balancing. Using cloud programming frameworks, application developers can focus on the functional aspects of their applications and benefit from the power of cloud computing. In this chapter, we will show how to use some of the existing cloud programming frameworks in three application domains: data analytics, Web applications, and distributed computing. More specifically, we will explain how to use MapReduce (Dean and Ghemawat, 2008) for data analytics, Google App Engine (Google, 2014) for Web applications, and SALSA (Varela and Agha, 2001) for distributed computing. The rest of the chapter is structured as follows. In section 50.2, we describe nonfunctional concerns supported at different levels of cloud services and go through existing cloud programming frameworks. In section 50.3, we explain MapReduce, Google App Engine, and Simple Actor Language System and Architecture (SALSA). In section 50.4, we illustrate how to use these three programming frameworks by showing example applications. Finally, we conclude the chapter in section 50.5. Developing Elastic Software for the Cloud

Extracted Key Phrases

Cite this paper

@inproceedings{Imai2016DevelopingES, title={Developing Elastic Software for the Cloud}, author={Shigeru Imai and Pratik Patel and Carlos A. Varela}, year={2016} }