Databases Reference
In-Depth Information
Chapter 5
Application Architectures
for Big Data and Analytics
Big data's bigness is hardly the interesting characteristic. The real fun lies
in how we think about data, where they reside in the data ecosystem and
how do we generate value from them.
Data-driven platforms are powering business innovations. Generally speaking,
transactional data created and stored by enterprise systems such as customer data
in CRM applications, operational data in ERP systems, financial data in accounting
databases, and sales and marketing data in sales and marketing applications constituted
the majority of business relevant data. This data was brought into
enterprise data
warehouse
(EDW) systems and BI applications for a consolidated enterprise-wide
view and business performance reporting. These system capabilities were gradually
challenged due to the scalability and performance considerations purely owing to
changing nature of data (volume, velocity, and variety).
Figure
5-1
illustrates the data spectrum along with the big data characteristics.
In the earlier chapters we have discussed the need for a radically different approach
to data processing and analytics to solve the challenges thrown by big data. In
the subsequent sections in this chapter we will discuss what different application
architectures we need to consider to implement big data and analytics solutions.
