Using semantic knowledge of transactions to increase concurrency
ACM Transactions on Database Systems (TODS)
Transaction chopping: algorithms and performance studies
ACM Transactions on Database Systems (TODS)
A concurrency control model for PDM systems
Computers in Industry
| Hi-index | 0.00 |
The concurrency control problem in database systems has been examined by many people and several concurrency control algorithms have been proposed. The most popular algorithms are two-phase locking and timestamp ordering. This paper shows that two-phase locking and timestamp ordering are special cases of a more general concurrency control algorithm. This general algorithm is described in detail and is proven to work correctly. We show that two-phase locking and timestamp ordering represent the two end points of a series of concurrency control algorithms. Each of them is a special case of the general algorithm proposed in this paper. Moreover, each of these special cases can be selected in advance, and can even be changed dynamically during execution.