Modern Big Data Architectures

CHAPTER 5Cloud Computing

In Chapter 2 we introduced cloud computing as one of major paradigms for building modern information systems. There is a strong rationale behind this trend. When on premise systems become overloaded and we enter the big data world, we need a means to scale with commodity hardware at hand or get external resources, which is what the Infrastructure as a Service (IaaS) cloud model provides.

Later in this chapter we will draw analogies between the cloud and multi-agent systems but also see how these technologies can work together in order to compose efficient platforms for information processing at scale.

5.1 Cloud Enabled Architectures

We will now dive deeper into cloud based architectures and the benefits they bring. In this chapter we will not distinguish transactional from analytical processing, assuming some abstract computational jobs are to be performed. More on specific computational models can be found in Chapter 6, where the IaaS cloud model can be applied. An in-depth review of big data analytics is located in Chapter 7. In the latter case also Platform as a Service (PaaS) and Software as a Service (SaaS) models can be applied if a cloud provider implements dedicated AI/ML capabilities.

5.1.1 Cloud Management Platforms

Regardless of the cloud model (Iaas, PaaS, SaaS, etc.), management of resources is not an easy task and requires specialized software called Cloud Management Platform (CMP). According to Gartner Gartner. CMPs are: Integrated ...