Data is being collected at an increasing rate and is relied on for decision making processes in a variety of industries. The modern day challenge of "big data" is occurring because many advancements in computer system performance occurred in the arena of CPU performance and massive parallelism. However, there is now a realization that I/O performance is lagging behind CPU performance in terms of performance gains since massive parallelism can be harnessed with cloud computing environments. Low I/O performance is now a limiting factor in the performance of large data analysis systems. However, increased I/O performance alone will not address all of the issues with analysis systems. This is because large scale data systems have placed many requirements on data collection, analysis and reporting systems. For example, performance, scalability and security are often key requirements for modern analysis systems. Security concerns and enterprise wide metadata are also necessary items to ensure that organizations are extracting the maximum benefit from centralized data warehouse and analysis systems. This work explores performance, scalability, security, and enterprise wide metadata in an effort to suggest how they can be jointly utilized to build scalable data analysis systems.