Abstract:This paper presents a benchmark for measuring the blocking behavior of schema transformations in relational database systems. As a basis for our benchmark, we have developed criteria for the functionality and performance of schema transformation mechanisms based on the characteristics of state of the art approaches. To address limitations of existing approaches, we assert that schema transformations must be composable while satisfying the ACID guarantees like regular database transactions. Additionally, we have identified important classes of basic and complex relational schema transformations that a schema transformation mechanism should be able to perform. Based on these transformations and our criteria, we have developed a benchmark that extends the standard TPC-C benchmark with schema transformations, which can be used to analyze the blocking behavior of schema transformations in database systems. The goal of the benchmark is not only to evaluate existing solutions for non-blocking schema transformations, but also to challenge the database community to find solutions that allow more complex transactional schema transformations.

In earlier work we have extended the TPC-C benchmark with basic and complex schema transformations. This paper uses this benchmark to investigate the blocking behaviour of online schema transformations in PostgreSQL, MySQL and Oracle 11g. First we discuss experiments using the data definition language of the DBMSs, which show that all complex operations are blocking, while we have mixed results for basic transformations. Second, we look at a technique for online schema transformations by Ronström, based on triggers. Our experiments show that pt-online-schema-change for MySQL and DBMS_REDEFINITION for Oracle can perform basic transformations without blocking, however, support for complex transformations is missing. To conclude, we provide a solution outline for complex non-blocking transformations.