OODBMS is widely favored for mapping domains like CAD, with complex representation requirements. The transactions accessing OODBMS can be categorized into runtime transactions and design time transactions. Runtime transactions are meant for accessing data and design time transactions are meant for accessing schema. Parallel execution of transactions is supported to promote concurrency and throughput. In continuously evolving business domains, frequent schema changes are done to reflect the changes of business domain. Then it implies that more number of design time transactions arrive parallely with runtime transactions. Concurrency control schemes are employed to maintain the consistency of the database. Several semantic multi-granular lock based concurrency control schemes have been proposed in the literature. They have the drawback of either poor performance or high maintenance overhead when applied to continuously evolving domains. This paper proposes semantic lock-based concurrency control mechanism with better performance and nil maintenance overhead for continuously evolving OODBMS. It uses lock rippling to improve the performance. © 2012 Springer-Verlag.
CITATION STYLE
Geetha, V., & Sreenath, N. (2012). High concurrency for continuously evolving OODBMS. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 7154 LNCS, pp. 94–105). https://doi.org/10.1007/978-3-642-28073-3_9
Mendeley helps you to discover research relevant for your work.