The Community for Technology Leaders
Green Image
<p><b>Abstract</b>—Performance analysis of concurrent executions in parallel systems has been recognized as a challenging problem. The aim of this research is to study approximate but efficient solution techniques for this problem. We model the structure of a parallel machine and the structure of the jobs executing on such a system. We investigate rich classes of jobs, which can be expressed by series, parallel-and, parallel-or, and probabilistic-fork. We propose an efficient performance prediction method for these classes of jobs running on a parallel environment which is modeled by a standard queueing network model. The proposed prediction method is computationally efficient, it has polynomial complexity in both time and space. The time complexity is <tmath>$O(C^{2}N^{2}K)$</tmath> and the space complexity is <tmath>$O(C^{2}N^{2}K)$</tmath>, where <tmath>$C$</tmath> is the number of job classes in the system, the number of tasks in each job class is <tmath>$O(N)$</tmath>, and <tmath>$K$</tmath> is the number of service centers in the queueing model. The accuracy of the approximate solution is validated via simulation.</p>
Concurrent programs, distributed systems, parallel processing, performance evaluation, queueing networks, task graphs.
De-Ron Liang, Satish K. Tripathi, "On Performance Prediction of Parallel Computations with Precedent Constraints", IEEE Transactions on Parallel & Distributed Systems, vol. 11, no. , pp. 491-508, May 2000, doi:10.1109/71.852402
82 ms
(Ver 3.3 (11022016))