高效能計算系統中資源碎片之利用

在高效能計算系統中，資源會根據一些排程方法被分配給使用者。但不幸的是，這些排程方法在分配的過程中通常會產生一些資源碎片，降低系統的資源使用率。在本論文中，我們提出了一個可以最大化系統資源使用率的高效能計算系統。這個系統藉由把沒有使用到的資源碎片分配給那些可以在任一時間點被開啟以及結束的可延展性工作，來讓系統的資源使用率接近100%。也因為這些工作是使用資源碎片來做計算，這些工作應該要以不同的價錢來作收費。

關鍵字

高效能計算；資源碎片；碎片

並列摘要

In high-performance computing (HPC) systems, resources are al-located to users according to configurable scheduling policies. Unfor-tunately, these policies often create resource fragments which reduce overall resource utilization. In this paper, we propose an HPC archi-tecture which maximizes the resource utilization by allocating unused resource fragments for jobs which are malleable in the sense that these jobs can be created and terminated at any time without severe penalties. Hence, the whole HPC system can utilize almost 100% of its computing resources by filling its schedule with these malleable jobs. Malleable job computation can then be offered at a lower price for potential cus-tomers, allowing for multi-pricing schemes for the HPC system pro-vider.

並列關鍵字

high-performance computing ； resource fragment ； fragment

參考文獻

[5] Anderson, David P.: BOINC: A system for public-resource com-puting and storage. In: Grid Computing, 2004, pp. 4-10. Pro-ceedings. Fifth IEEE/ACM International Workshop, IEEE. (2004)

[9] I-Chen Wu, Hung-Hsuan Lin, Der-Johng Sun, et al.: Job-Level Proof Number Search. In: the IEEE Transactions on Computational Intel-ligence and AI in Games (SCI), Vol. 5, No. 1, pp. 44-56, March (2013)

[14] Zhou, Songnian.: LSF: Load Sharing in Large Heterogeneous Dis-tributed Systems. In: Workshop on Cluster Computing. (1992)

[1] Kalé, Laxmikant V., Sameer Kumar, and Jayant DeSouza.: A malleable-job system for timeshared parallel machines. In: Cluster Computing and the Grid, 2002, pp. 230-230. 2nd IEEE/ACM In-ternational Symposium, IEEE (2002)

Google Scholar

[2] Microsoft Azure, http://azure.microsoft.com/

Google Scholar

國際替代計量

高效能計算系統中資源碎片之利用

全文下載

主題瀏覽