Authors:
Ming Jiang
1
;
Tom Kirkham
2
and
Craig Sheridan
3
Affiliations:
1
University of Sunderland, United Kingdom
;
2
Science and Technology Facilities Council, United Kingdom
;
3
Flexiant Limited, United Kingdom
Keyword(s):
Cultural Algorithm, Service Reliability, Risk Management, Virtual Machine Scheduling, Optimisation.
Related
Ontology
Subjects/Areas/Topics:
Cloud Computing
;
Cloud Middleware Frameworks
;
Cloud Risk, Challenges, and Governance
;
Fundamentals
;
Platforms and Applications
;
Services Science
;
Services Security and Reliability
Abstract:
Cloud service reliability is one of the key common performance concerns of both Cloud Service Provider (CSP) and Cloud Service User (CSU). As the capability and scale of a Cloud infrastructure increase, the requirements of maintaining and improving the reliability of services is increasingly crucial for the CSP and CSU. Risk management is the process of analysing the potential risk factors associated with the reliability deterioration of a service provided by a CSP, assessing the uncertainties and consequences associated with this kind of deterioration, and finally identifying the system wide appropriate mitigation strategies for risk treatments. In this paper, an evolutionary Cultural Algorithm based risk management method is proposed to facilitate the identification (i.e., probability and consequences) and treatment (i.e., mitigations) of Cloud infrastructure reliability related risk for Virtual Machine scheduling optimisation.