Business opportunity exists in leveraging dynamic, rich data, often ‘big data’, in real-time in today's operational systems. Rich data may present a clear business opportunity, but slow queries will just as quickly wipe it out. Customers don’t have time for your database. A time-to-interaction of more than 1 second will start losing your customers, after 15 seconds they will all be gone.
Working with its customers, Akiban has surveyed over 50,000 problematic queries in MySQL applications. We have converged on the most frequently encountered problem query patterns and then illustrated how Akiban’s table-grouping accelerates those queries in real world scenarios:
| Business Need | Technical Challenge |
| User Customization Customization of applications such as CMS, CRM or eCommerce, through the addition of custom fields |
User-driven addition of tables and columns, leading to over-normalized schema and poor indexing (more ..) |
| Dynamic Profile Search Search on two or more criteria that change in real-time and are from different, large tables, presented in order |
Predicates on multiple tables, leading to creation of temporary tables that need to be sorted (more ..) |
| Social Context Presentation of constantly changing, aggregated information within groups |
Tables need to be joined into temp tables, then sorted and scanned for distinct values to be aggregated and grouped (more ..) |
| Code-Generated SQL Using ORMs or form-based SQL generators to hide the complexity of databases from users or programmers |
Subqueries that are executed for every row that is returned from the master query. (more ..) |
