PIPELINED PARALLELISM SCHEDULING IN HOMOGENEOUS AND HETEROGENEOUS ENVIRONMENTS

Pipelined parallelism scheduling problem is very important in the area of parallel query optimization. The scheduling problem is modeled by using a Pipelined Operator Tree (POT), which is a tree whose nodes represent query operators that can be run in parallel and edges represent communication between adjacent operators; we must find a schedule for the POT that minimizes the total response time, a problem which has been shown to be NP-hard. This paper presents algorithms (some of them proposed by authors) for pipelined parallelism scheduling for homogeneous and heterogeneous environments and compares their by computer-aided analysis, simulating their behaviors.