近年来, 云计算已经成为一种支持按需(on-demand)提供计算资源的低成本传输模式. 在云平台上, 弹性的资源使用是一种基于“现用现付”(pay-as-you-go)的商业模式, 通过“按需”的原则来提供弹性的资源. 介绍一种新的弹性算法(elastic algorithm, EA), 即算法本身就是通过“现用现付”的方式组织起来. 在传统算法中, 计算是一个确定过程, 只会产生一种完整的结果或者没有结果. 与之对比, 弹性算法会随着资源的消耗而生成一组近似结果. 具体来说, 随着消耗的资源越来越多, 弹性算法能够保证产生更好质量的结果. 在这个意义上, 算法产生结果质量的好坏依赖于资源消耗的多少, 因此是具有弹性的. 最后, 正式定义弹性算法的必要性质, 并对弹性算法未来的研究方向进行展望, 提出一系列的研究挑战.
In recent years, cloud computing has emerged as a cost-effective way to deliver on-demand and metered computing resources. In a cloud, elasticity of resource usage is typically realized through the “on-demand” provision principle supported by the “pay-as-you-go” business model. However, little has been investigated into elasticity of algorithm for cloud computing. This paper introduces a novel elastic algorithm (EA) in which the computation itself is organized in a “pay-as-you-go” fashion. In contrast to conventional algorithms, where computation is a deterministic process that only produces an “all-or-nothing” result, an EA generates a set of approximate results corresponding to its resource consumption. As more resources are consumed, better results can be derived. In this sense, quality of the algorithm is elastic to its resource consumption. The desirable properties for EA are formalized, and ambitious agenda for future research is provided in this area and propose several challenges.
[1] Han R, Guo L, Guo Y, et al. A deployment platform for dynamically scaling applications in the cloud [C]// The 3rd IEEE International Conference on Cloud Computing
Technology and Science. 2011: 506-510.
[2] Han R, Ghanem M M, Guo L, et al. Enabling costaware and adaptive elasticity of multi-tier cloud applications [J]. Future Generation Computer Systems, 2012, DOI: 10.1016/j.future.2012.05.018.
[3] Han R, Guo L, Ghanem M M, et al. Lightweight resource scaling for cloud applications [C]// The 12th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing. 2012: 644-651.
[4] Poladian V, Sousa J P, Garlan D, et al. Dynamic configuration of resource-aware services [C]// The 26th International Conference on Software Engineering. 2004: 604-613.
[5] Gaber M M, Yu P S. A framework for resourceaware knowledge discovery in data streams: a holistic approach with its application to clustering [C]// The 2006 ACM Symposium on Applied Computing. 2006: 649-656.
[6] Dean T L. Intractability and time-dependent planning [C]// The 7th National Conference on Artificial Intelligence. 1989: 245-266.
[7] Dean T, Boddy M. An analysis of time-dependent planning [C]// The 7th National Conference on Artificial Intelligence. 1988: 49-54.
[8] Zilberstein S. Using anytime algorithms in intelligent systems [J]. AI Magazine, 1996, 17: 73-83.
[9] Guo Y, Ghanem M, Han R. Does the cloud need new algorithms: an introduction to elastic algorithms [C]// The 4th International Conference on Cloud Computing. 2012: 66-73.
[10] Pharr M, Humphreys G. Physically based rendering: from theory to implementation [M]. San Fransisco: Morgan Kaufmann, 2004: 1-43.
