Databases Reference
In-Depth Information
integrated architecture that is deployed in this enterprise is the actual footprint that will begin to exist
as the Big Data - data warehouse.
Hadoop and MySQL drives innovation
Another implementation of a Big Data-based data warehouse architecture is driven by integrat-
ing and augmenting the incumbent platform with a Hadoop and MySQL architecture. The difference
in this architecture approach is the migration from a traditional RDBMS to the new data architecture
platform.
The business problem is a leading electronics manufacturer has been having a weak market penetra-
tion with its products and services. The biggest threats facing the enterprise include loss of traditional
markets, increased customer attrition, poor market performance from a wallet-share perspective, lack of
customer confidence, and overall weak performance.
The executive teams within the enterprise were unhappy about the situation and were mandated by
the board to regain the brand value in the market. The team started conducting studies to understand the
cause of the issues, and found the following:
●
The product suite manufactured by the enterprise was not reflecting the market demands of the
geographies they were sold to.
●
The competition was producing similar products with slightly better features at a lower cost.
●
The services provided by the enterprise after the product sale were minimal and lackluster.
●
The market pricing strategy did not accurately reflect socioeconomic conditions.
●
Customer reviews and feedback about the products and services were largely not considered
beyond market research teams.
●
Call center teams were not equipped with the right information when customers called for help
and advice, resulting in disappointing and embarrassing situations.
Deeper investigation into the current-state issues revealed the following architectural issues:
●
The source systems were deployed globally and the data in each of these systems was based on
requirements for the local market.
●
Each source system had different pieces of information about the product.
●
All the product data was buried in textual manuals and was not available digitally.
●
Each region of the world had customer feedback both solicited via surveys and unsolicited in web
forums and social media.
●
Competitive research teams across the world were working on different data sets.
●
Vendor management was not a centralized function.
●
Financial systems were consolidating data from multiple systems across the world. The
integration was not based on auditable data.
The enterprise architecture and data teams were assigned with the task of creating a future-state
architecture that enabled the incumbent data warehouse to be extensible and scalable, while meeting
the requirements of the enterprise globally. The teams decided to evaluate the following platforms in
addition to the incumbent RDBMS platform:
●
Hadoop—the Big Data platform can be deployed as the enterprise repository of data. With its
lower cost and greater scalability, the platform will bring a lot of performance boost to the overall
data architecture.
