Primary copy synchronization for DB-sharing
Information Systems
The effect of skewed data access on buffer hits and data contention in a data sharing environment
Proceedings of the sixteenth international conference on Very large databases
An approximate analysis of the LRU and FIFO buffer replacement schemes
SIGMETRICS '90 Proceedings of the 1990 ACM SIGMETRICS conference on Measurement and modeling of computer systems
Performance analysis of coherency control policies through lock retention
SIGMOD '92 Proceedings of the 1992 ACM SIGMOD international conference on Management of data
Analysis of superposition of streams into a cache buffer
SIGMETRICS '93 Proceedings of the 1993 ACM SIGMETRICS conference on Measurement and modeling of computer systems
VAXcluster: a closely-coupled distributed system
ACM Transactions on Computer Systems (TOCS)
Some Distribution-Free Aspects of Paging Algorithm Performance
Journal of the ACM (JACM)
Stochastic Analysis of Computer Storage
Stochastic Analysis of Computer Storage
IEEE Transactions on Computers
Performance Analysis of Affinity Clustering on Transaction Processing Coupling Architecture
IEEE Transactions on Knowledge and Data Engineering
Performance Analysis of Buffer Coherency Policies in a Multisystem Data Sharing Environment
IEEE Transactions on Parallel and Distributed Systems
Performance Evaluation of Transaction Processing Coupling Architectures for Handling System Dynamics
IEEE Transactions on Parallel and Distributed Systems
Efficient Locking and Caching of Data in the Multisystem Shard Disks Transaction Environment
EDBT '92 Proceedings of the 3rd International Conference on Extending Database Technology: Advances in Database Technology
Database Buffer Model for the Data Sharing Environment
Proceedings of the Sixth International Conference on Data Engineering
Effect of System Dynamics on Coupling Architectures for Transaction Processing
Proceedings of the Eighth International Conference on Data Engineering
A Simple Analysis of the LRU Buffer Policy and Its Relationship to Buffer Warm-Up Transient
Proceedings of the Ninth International Conference on Data Engineering
| Hi-index | 0.00 |
The performance of a transaction processing system is very sensitive to the buffer hit probability. In a data sharing environment where multiple computing nodes are coupled together with direct access to shared data on disks, buffer coherency needs to be maintained such that if a data granule is updated by a node, the old copies of this granule present in the buffer of other nodes must be invalidated. The buffer invalidation phenomenon reduces the buffer hit probability in a multi-node environment. After the buffer reaches a certain size, the buffer hit probability will remain constant regardless of further increase in buffer size due to the buffer invalidation effect. This puts an upper limit on the achievable buffer hit probability. Thus the selection of appropriate buffer size is one of the critical issues in a data sharing environment. In this paper, we develop an asymptotic analysis of the Markov model for a buffer in the data sharing environment. Important relations between buffer size, number of nodes, write-probability and the size of the database to the buffer hit probability had been found in all range of system parameters. A simple expression is obtained for the maximum achievable buffer hit probability and also for the maximum usable buffer size. Various properties of the maximum achievable buffer hit probability and usable buffer size are derived for a skewed access workload. The accuracy of the asymptotic method is validated by numerous case studies.