Mechanisms for Parallel Query Execution

Parallel query processing has attained a lot of attention in the database community because customer applications on databases are growing in size and complexity with an ever increasing demand for performance. The main objective of parallel query processing is to achieve speedup, scaleup and high throughput. Earlier bracket model was used for parallelization. In case of bracket model it is not possible to possible to execute two operators in one process without resorting to some thread or coroutine facility. Also bracket model involves expensive interprocess communication system calls even in the cases when an entire query is evaluated on a single CPU. To overcome all these difficulties of bracket model we have developed a new model in which a single operator is used for parallelization. As compared to bracket model, all issues of control are localized in one operator (parallelizing operator) in this model. Also we can execute a complex query in a single or with a number of processes by using one or more parallelism operators in the query evaluation plan.