We are going through a period of rapid development in data management systems. On one side of the spectrum, commercial database system vendors are adding new features related to ease of management, semistructured data, data compression, parallelism, reliability, and security. On the other side, new data-processing systems are being developed over MapReduce backends, Key-Value stores, low-power devices, and widely-distributed systems.

New usage patterns, evolving hardware trends, and increased competition drive continuous innovation and expansion of these systems; both in terms of features as well as code size. As a result, it is increasingly expensive to test and tune these systems, and these stages tend to dominate the release cycle. It is not unusual to find fifty percent of the development cost being spent on testing and tuning, and several months being reserved for testing before a new release can be shipped. This situation will get worse in the future unless revolutionary new ideas are brought in.

The first two workshops on testing database systems (collocated with SIGMOD 2008 and 2009) have shown that there is a huge interest in the industry to discuss problems in the area of testing database systems together with the academic community. Testing has gained more attention in the database community recently with an increasing number of conference submissions as well as a special issue of the IEEE Data Engineering Bulletin in this area. DBTest 2010 will continue the discussion between industry and academia in order to come up with a research agenda that describes important open problems in the area of testing database systems and applications. The long-term goal is to devise new techniques that solve these problems. These techniques will reduce the cost and time to test and tune database products so that users and vendors can spend more time and energy on actual innovations.
Topics of Interest

* Testing techniques for database systems, data storage services, and database applications
* Generation of synthetic data for test databases
* Generation of stochastic models for large test matrices
* Algorithms for automatic program verification
* Maximizing code coverage during testing of database systems and applications
* Testing the reliability and availability of database systems
* Testing and designing systems that are robust to estimation inaccuracies
* Testing the effectiveness of adaptive policies and components
* Interactions between testing and tuning of database systems
* Identifying performance bottlenecks
* Workload characterization with respect to performance metrics and engine components
* Testing data consistency Vs. availability tradeoffs in emerging systems
* Metrics for predictability of query and workload performance
* Metrics for query plan robustness
* Security and vulnerability testing
* War stories and vision papers