The Challenges in Deploying a Data Warehouse and How to Overcome Them
October 24, 2013

Data is becoming increasingly important to businesses. Companies need up-to-date data as well as need it to be easily accessible so that they can make informed business decisions. From sales forecasts to trending patterns of users and customers—and everything in between—data is fast becoming one of the most intrinsic pieces of intellectual property in companies. In the world of data, solutions built to deliver this information are known as data warehouses.

A data warehouse solution is often not the same as a traditional relational database management system (RDBMS). In many cases, data warehouse solutions are specially architected from ground up and deployed on dedicated hardware
or in a dedicated virtual machine. Having a data warehouse implementation that is separate from a core read/write RDBMS includes the following benefits:

  • Data from multiple source systems can be easily consumed, rolled up, transformed, and aggregated in a single system. This allows for easier reporting and analysis, including increased performance.
  • Data quality is improved because the data is cleaned on the way into the warehouse. Improved data quality allows business users to have more accurate answers and provides a single view of “the truth”.
  • A data warehouse schema is usually optimized for analysis. Many times, the source system's schema is not optimal for analysis-based questions. This is not to say a query cannot be run, but in a non-optimized source, the same query may consume a higher amount of resources, take a longer time to run, and not provide the same analysis capabilities.
  • A data warehouse allows for easier retention of business critical data, even if the data does not need to exist in the core read/write database. Since the data required is usually a subset of what is in the main database, while it will consume pace over time, it would be less than the space needed if all the data was kept in the individual sources.

All of this ties into the concept of big data.

Although a data warehouse provides many advantages, there are always challenges related to its deployment, including:

  • Performance and scalability
  • Dealing with structured versus unstructured data
  • Reliability, manageability, and support
  • Cost

This paper will discuss those challenges and how the HP AppSystem for Microsoft SQL
Server 2012 Parallel Data Warehouse (PDW) appliance can help you overcome them.

Download it now>>