Provisio, provider of the clinical trial volunteer-finder service iTrials, has hit pay dirt in a column-based database architecture. The highly data-intensive processing required for its service had proven too much for its more traditional database.
The company's new hardware/software implementation required a data center upgrade and some customization of internal applications. But they yielded a reduction in database tables to just 12, down from the original 230,000. Queries that required 10 minutes of processing are now complete in 30 seconds, according to Search Data Management.
The service for finding clinical trial subjects involves an automated search for a set of people with a specific illness, and then winnowing the records to find those who meet specific criteria. Provisio says its row-based relational database had hit its limit. Queries were taking longer than usual to run and the process was plagued with performance bottlenecks. It opted for the column-based database to "get to the data faster."
- see the article