SUMMARY
This paper discusses the selection of the access path in the relational database management system. It is described in terms of single table queries. The right choice, which is based on the evaluation work, are compared and discussed. The results, which have been made in the results, shows that the cost which is forecasted through the optimizers are seen that they are not so true in the absolute value. So in the majority of the cases, the true optimal path is selected. In most of the cases, the ordering, which is made among the estimated cost for the entire path, is seen to be the same, which is the actual measured cost. Moreover, the cost, which has occurred in the selection of the path, is also not great. The cost of the optimization in a two-way join is seen to be equivalent between the 5 to the 20 database retrievals. It is seen that this number even become more confusing when the selector of the path is placed in the system R environment in which the application programs are supposed to comply once and they are run many times. The optimization cost is thought to be amortized and run many times.
The important role of this path selection is the expanded use of statistics over any other area. it can be it can be index cardinality .the CPU utilization can also be included in the cost formulas as well as the join order method for the purpose of determination. There are many queries, which are related to the CPU-bound especially related to the merge joins for which the temporary relations are made as well as performed. There is also the “selectivity factor” concept, which gives permission to the optimizers that they can take advantage of many quest restriction predicates in the research of RSS. In the interesting ordering, there are equivalent classes for the order and join and group specification. The optimizers are performing more bookkeeping than the selectors but this often results in avoiding the sorting as well as storage of the results of an intermediate query. The techniques relate to tree searching and tree pruning gives permission to d other additional bookkeeping in order to perform effectively. There is a need to more work done on the validation of the optimizer cost but it can be concluded that database management system is able t support the nonprocedural query languages with the performance that are comparable to those that are supporting the current languages which are more procedural related to the database management system.
Opinion:
This paper is based on the data manipulation language done in a high-level query; the languages like the SQL are selected. The requests made in this paper are non-procedurally stated and these are done without reference to the access path. In the opinion, this paper perfectly describes that how the system R selects the access path for the simple in which there is a single relation as well as complex queries, which includes the joints. The paper describes the user specification of the desired data, which is a Boolean, expression done in the sense of predicates. In this paper, the system R us seen as the experimental database management system which is developed for the purpose of carrying out the research on the data of the relational model.
This is done perfectly because of the fact that the system R was designed as well as built through the help of the IBM members in the research laboratory. The computation, interpretation made in this paper is done through the proper methods and the results are proved reliable because they are performed under the valid method, which proves that this research is of high quality.