Abstract

Despite the huge success in various worldwide projects, volunteer computing also suffers from the possible lack of computing resources (one volunteered device can join one project at a time) and from the uncertain job interruptions (the volunteered device can crash or disconnect from the Internet at any time). To relieve the challenges faced by volunteer computing, we have proposed bonus computing that exploits the free quotas of public Cloud resources particularly to deal with problems composed of fine-grained, short-running, and compute-intensive tasks. In addition to explaining the loosely-coupled functional architecture and six architectural patterns of bonus computing in this paper, we also employ the Monte-Carlo approximation of Pi (π) as a use case demonstration both to facilitate understanding and to help validate its functioning mechanism. The results exhibit not only effectiveness but also multiple advantages of bonus computing, which makes it a valuable evolution from and supplement to volunteer computing.

Recommended Citation

Li, Z., Chen, Y., Rodríguez, M., & Deng, L. (2018). Bonus Computing: An Evolution from and a Supplement to Volunteer Computing. In B. Andersson, B. Johansson, S. Carlsson, C. Barry, M. Lang, H. Linger, & C. Schneider (Eds.), Designing Digitalization (ISD2018 Proceedings). Lund, Sweden: Lund University. ISBN: 978-91-7753-876-9. http://aisel.aisnet.org/isd2014/proceedings2018/ISDevelopment/3.

Paper Type

Event

Share

COinS
 

Bonus Computing: An Evolution from and a Supplement to Volunteer Computing

Despite the huge success in various worldwide projects, volunteer computing also suffers from the possible lack of computing resources (one volunteered device can join one project at a time) and from the uncertain job interruptions (the volunteered device can crash or disconnect from the Internet at any time). To relieve the challenges faced by volunteer computing, we have proposed bonus computing that exploits the free quotas of public Cloud resources particularly to deal with problems composed of fine-grained, short-running, and compute-intensive tasks. In addition to explaining the loosely-coupled functional architecture and six architectural patterns of bonus computing in this paper, we also employ the Monte-Carlo approximation of Pi (π) as a use case demonstration both to facilitate understanding and to help validate its functioning mechanism. The results exhibit not only effectiveness but also multiple advantages of bonus computing, which makes it a valuable evolution from and supplement to volunteer computing.