Databases Reference
In-Depth Information
DSS 2.0 is implemented more from the front-end solution with common metadata integration
points in the industry today.
The challenge that faces data warehousing today is not the volume of data alone; the data types
and data formats are issues that threaten to violate the data processing rules of the data warehouse.
The rules that were designed for relational data cannot be enforced on text, images, video, machine-
generated data, and sensor data. The future data warehouse will also need to handle complex event
processing and streaming data and support object-oriented or Service Oriented Architecture (SOA)
structures as part of data processing.
The future evolution of data warehousing will be an integration of different data types and their
usage, which will be measured as the workload executed on the data warehouse. The next generation
of the data warehouse design technique will be the workload-driven data warehouse, where the funda-
mental definition of the data warehouse remains as defined by the founding fathers, but the architec-
ture and the structure of this data warehouse transcends into heterogeneous combinations of data and
infrastructure architectures that are based on the underlying workloads.
While Inmon's DW 2.0 and DSS 2.0 architectures provide foundational platforms and approaches
for the next-generation data warehouse, they focus on usability and scalability from a user perspec-
tive. The workload-driven data warehousing architecture is based on functionality and infrastructure
scalability, where we can compartmentalize the workloads into discrete architectures. In a sense, the
workload-driven architecture can be considered as a combination of Inmon's DW 2.0 and Google's
Spanner architecture.
SUMMARY
As we conclude this chapter, readers need to understand the primary goals of a data warehouse
will never change, and the need to have an enterprise repository of truth will remain as a constant.
However, the advances in technology and the commoditization of infrastructure now provide the
opportunity to build systems that can integrate all the data in the enterprise in a true sense, provid-
ing a holistic view of the behavior of the business, its customers, and competition, and much more
decision-making insights to the business users.
Chapter 7 will focus on providing you with more details regarding workload-driven architecture
and the fundamental challenges that can be addressed in this architecture along with the scalability
and extensibility benefits of this approach.
Further reading
