Enhancing pre-existing data managers with atomicity and durability
Many of currently available managers of persistent data do not provide support for updates that are atomic and durable. Developing from scratch the requisite software for incorporating the atomicity and durability properties into each of these pre-existing data managers is a time-consuming and formidable task. It is hence desirable to design and build a general purpose recovery manager that can be easily interfaced with pre-existing data managers. In this paper, we consider four recovery manager architectures that are based on recovery algorithms proposed in the literature and recovery designs adopted in existing products. We evaluate the various architectures with respect to 1) the extent of the modifications they cause to the software of pre-existing data managers, and 2) the degree to which they affect the performance of the data managers. Finally, we relate the experience we had in integrating a spatial data manager with a general purpose recovery manager.
Unable to display preview. Download preview PDF.
- BHG87]P. A. Bernstein, V. Hadzilacos, and N. Goodman. Concurrency Control and Recovery in Database Systems. Addison-Wesley, Reading, MA, 1987.Google Scholar
- [GR91]J. Gray and A. Reuter. Transaction Processing: Concepts and Techniques. Morgan Kaufmann, San Mateo, California, 1991.Google Scholar
- [Lom92]D. Lomet. MLR: A recovery method for multi-level systems. In Proceedings of A CM-SIGMOD 1992 International Conference on Management of Data, San Diego, California, pages 185–194, 1992.Google Scholar
- [MHL+92]C. Mohan, D. Haderle, B. Lindsay, H. Pirahesh, and P. Schwarz. ARIES: A transaction recovery method supporting fine-granularity locking and partial rollbacks using write-ahead logging. ACM Transactions on Database Systems, 17(1):94–162, March 1992.Google Scholar
- [RKS93]R. Rastogi, H. F. Korth, and A. Silberschatz. Strict histories in objectbased database systems. In Proceedings of the Twelfth ACM SIGACT-SIGMOD-SIGART Symposium on Principles of Database Systems, Washington D.C., 1993.Google Scholar
- [RN92a]R. Rastogi and M. A. Netmat. Enhancing pre-existing recovery managers with atomicity and durability. Technical Report 113880-940420-24, AT&T Bell Laboratories, 1992.Google Scholar
- [RN92b]R. Rastogi and M. A. Neimat. On using the encina toolkit to provide recovery support for data managers. Technical Report HPL-DTD-92-21, Hewlett-Packard Laboratories, Palo Alto, 1992.Google Scholar
- [Tra]Transarc Corporation. Encina Toolkit Server Core: Programmer's Reference.Google Scholar
- [WHBM90]G. Weikum, C. Hasse, P. Broessler, and P. Muth. Multi-level recovery. In Proceedings of the Nineth ACM SIGACT-SIGMOD-SIGART Symposium on Principles of Database Systems, Nashville, pages 109–123, 1990.Google Scholar