Transaction processing; Transaction scheduling
Transaction management [1, 2] refers to the tasks of processing multiple transactions issued by various clients of a database server in such a way that the ACID contract can be fulfilled, that is, the properties of atomicity, consistency preservation, isolation, and durabilityof each individual transaction can be guaranteed. Transaction management is generally understood as requiring serializability-based concurrency control as well as recovery from failures. Concurrency control is the task of scheduling transactions such that their serializability can be guaranteed, while recovery has to restore a consistent database state after a system or media failure. Assuming that the database server is in charge of the “C,” the former guarantees the “I” in ACID, the latter the “A” and “D” properties. Transaction management has to be highly efficient, as modern transaction servers need to accommodate thousands of transactions...
- 3.Gray J, Lorie RA, Putzolu GR, Traiger IL. Granularity of locks in a large shared data base. In: Proceedings of the 1st International Conference on Very Data Bases; 1975. p. 428–51.Google Scholar
- 4.Bernstein PA, Hadzilacos V, Goodman N. Concurrency control and recovery in database systems. Reading: Addison-Wesley; 1987.Google Scholar
- 7.Weikum G, Vossen G. Transactional information systems – theory, algorithms, and the practice of concurrency control and recovery. San Francisco: Morgan Kaufmann; 2002.Google Scholar
- 8.Agrawal D, Das S, El Abbadi A. Data management in the cloud – challenges and opportunities. London: Morgan & Claypool Publishers; 2013.Google Scholar
- 10.Shasha D, Bonnet P. Database tuning: principles experiments and troubleshooting techniques. San Francisco: Morgan Kaufmann; 2002.Google Scholar