SLA Based Dynamic Provisioning of Cloud Resource in OLTP Systems

In the era of cloud computing, an increasing amount of services are moving online. Also increasing is the amount of cloud resource to power these services. Among these modern online transactions, many belong to the category of Online Transaction Processing (OLTP), which can be processed with predictable time and resource. However, with a large user base and fluctuated usage patterns, providing OLTP services efficiently remains a major challenge. In this paper we present an online algorithm that solves for a cost-minimizing provision scheme under fluctuated user requests, constrained by a tail-distribution-based Service Level Agreement (SLA), and incorporated with Neural Network prediction. Experiment shows that the algorithm delivers significant savings in provision, and outperforms a simple look-forward provision plan with the same SLA compliance.

[1]  Changtao Qu,et al.  Towards Autonomous Brokered SLA Negotiation , 2006 .

[2]  Rajkumar Buyya,et al.  Integrated Risk Analysis for a Commercial Computing Service , 2007, IPDPS.

[3]  Gianfranco Ciardo,et al.  Profit-driven service differentiation in transient environments , 2003, 11th IEEE/ACM International Symposium on Modeling, Analysis and Simulation of Computer Telecommunications Systems, 2003. MASCOTS 2003..

[4]  Muthucumaru Maheswaran,et al.  Utilizing unreliable public resources for higher profit and better SLA compliance in computing utilities , 2006, J. Parallel Distributed Comput..

[5]  Akshat Verma,et al.  On admission control for profit maximization of networked service providers , 2003, WWW '03.

[6]  Bill Claybrook.  OLTP: Online Transaction Processing Systems , 1992 .

[7]  Isi Mitrani,et al.  Management of Server Farms for Performance and Profit , 2010, Comput. J..

[8]  Dirk Neumann,et al.  A Service Request Acceptance Model for Revenue Optimization - Evaluating Policies Using a Web Based Resource Management Game , 2010, 2010 43rd Hawaii International Conference on System Sciences.

[9]  Danilo Ardagna,et al.  SLA based profit optimization in autonomic computing systems , 2004, ICSOC '04.

[10]  Yun Chi,et al.  SLA-Aware Profit Optimization in Cloud Services via Resource Scheduling , 2010, 2010 6th World Congress on Services.

[11]  Susanne Strahringer,et al.  Dynamically Scalable Architectures for E-Commerce , 2010, MKWI.

[12]  Raouf Boutaba,et al.  On Leveraging Policy-Based Management for Maximizing Business Profit , 2007, IEEE Transactions on Network and Service Management.

[13]  Asser N. Tantawi,et al.  Dynamic Application Placement Under Service and Memory Constraints , 2005, WEA.

[14]  Xing Wang,et al.  Profit and penalty aware (PP-aware) scheduling for tasks with variable task execution time , 2010, SAC '10.

[15]  Rada Chirkova,et al.  Dynamic Request Allocation and Scheduling for Context Aware Applications Subject to a Percentile Response Time SLA in a Distributed Cloud , 2010, 2010 IEEE Second International Conference on Cloud Computing Technology and Science.

[16]  Thomas F. Wenisch,et al.  PowerNap: eliminating server idle power , 2009, ASPLOS.