A Reinforcement Learning Based Workflow Application Scheduling Approach in Dynamic Cloud Environment

Workflow technology is an efficient means for constructing complex applications which involve multiple applications with different functions. In recent years, with the rapid development of cloud computing, deploying such workflow applications in cloud environment is becoming increasingly popular in many fields, such as scientific computing, big data analysis, collaborative design and manufacturing. In this context, how to schedule cloud-based workflow applications using heterogeneous and changing cloud resources is a formidable challenge. In this paper, we regard the service composition problem as a sequential decision making process and solve it by means of reinforcement learning. The experimental results demonstrate that our approach can find near-optimal solutions through continuous learning in the dynamic cloud market.

[1]  Enda Barrett,et al.  Applying reinforcement learning towards automating resource allocation and application scalability in the cloud , 2013, Concurr. Comput. Pract. Exp..

[2]  Bin Zhang,et al.  An Adaptive Decision Making Approach Based on Reinforcement Learning for Self-Managed Cloud Applications , 2016, 2016 IEEE International Conference on Web Services (ICWS).

[3]  Shijun Liu,et al.  A cost-optimal service selection approach for collaborative workflow execution in clouds , 2016, 2016 IEEE 20th International Conference on Computer Supported Cooperative Work in Design (CSCWD).

[4]  Bernd Kleinjohann,et al.  Learning Recommendation System for Automated Service Composition , 2013, 2013 IEEE International Conference on Services Computing.

[5]  Peter Dayan,et al.  Technical Note: Q-Learning , 2004, Machine Learning.

[6]  Xiang Zhou,et al.  Adaptive Service Composition Based on Reinforcement Learning , 2010, ICSOC.

[7]  Fairouz Fakhfakh,et al.  A Provisioning Approach of Cloud Resources for Dynamic Workflows , 2015, 2015 IEEE 8th International Conference on Cloud Computing.

[8]  Dazhong Wu,et al.  Cloud-based design and manufacturing: A new paradigm in digital manufacturing and design innovation , 2015, Comput. Aided Des..

[9]  Andrew W. Moore,et al.  Reinforcement Learning: A Survey , 1996, J. Artif. Intell. Res..

[10]  Mohammad Masdari,et al.  Towards workflow scheduling in cloud computing: A comprehensive analysis , 2016, J. Netw. Comput. Appl..

[11]  Sucha Smanchat,et al.  Taxonomies of workflow scheduling problem and techniques in the cloud , 2015, Future Gener. Comput. Syst..