Task Scheduling And Multiple Sequence Alignment In Volunteer Computing Platform P2HP

As the increasing of the number of high perform computing issues, traditional high perform computer is not satisfying in expansibility and cost. The network based distributed computing platform can resolve some of these problems. However, most of network based computing platforms at present are special platforms. Aimed at the problems of which, an open network based computing platform (P2HP) is constructed. P2HP not only adopt the mechanism of separation between the application system and core software, but also provide a convenient user developmental kit. Then users can design their applications by the kit.The task scheduling system is one of the central subsystems of P2HP. Task scheduling system of P2HP is not only required to be high efficiency and expansible, but also is demanded to be fault-tolerance and load-balance. To achieve these, we bring forward a Two Layer United Scheduling Algorithm (TLUSA) as the core algorithm of the task scheduling system of P2HP. Load-balance, idle computational ability of volunteer and correlative requirement of subtasks is taken into account in the algorithm, TLUSA. It decentralizes scheduling-load, avoids scheduling bottleneck, and improves dynamic expansibility and computing efficiency of P2HP. Some technologies have been adopted by task fault tolerance mechanism of TLUSA, such as overtime checking, redundancy computing. Fault tolerance mechanism has been adopted in the processing of communication. The task scheduling can be more stable via overtime checking and message re-passing. The D_ClustalW algorithm is a kind of P2HP based distributed multiple sequence alignment algorithm. The scheduling feature of P2HP and computational specialty of ClustalW algorithm have been taken into account by D_ClustalW. There are two kinds of task partition strategy in D_ClustalW: symmetrical partition and time forecast based partition.Finally, we give a performance evaluation by the simulative test and practical test. We align fifty SARS coronavirus genomes in the test. A personal computer cost more than 90 hours in full speed when align the same. The Speedup of P2HP is 5.7 when there are 8 volunteers in the P2HP. When more volunteers join, the speedup of platform can increase constantly. The test reveals that the P2HP is a high efficient platform and the task scheduling algorithm of P2HP is high efficient algorithm.