4
CREATING A MODERN ANALY TICS ARCHITECTURE
analytics
The challenge with existing
data infrastructures
Almost all organizations have built data warehouses for reporting and analytics
purposes. They use data from a variety of sources, including their own
transaction-processing systems and other databases. Many have also built
Hadoop frameworks for analyzing what is commonly called "big data" or data
that does not fit well in highly structured data warehouses. Building and running a
data warehouse and a big data framework have been complicated and expensive.
Traditional data warehouse challenges
Traditional data warehousing systems create a range of issues and demands:
• Cost millions of dollars in upfront software and hardware expenses
• Take months in planning and procurement
• Difficult to set up
• Need time for implementation and deployment processes
• Require that you define your data models and ingest data
• Hire a team of data warehouse administrators
• Keep your queries running fast and protect against data loss
• Only highly normalized data needed for mission-critical analytics
• A lot of data (dark data) in many siloed data stores
• Dark data never makes it into a data warehouse for analysis
• Difficult to scale
