Towards Risk-Aware Cost-Optimal Resource Allocation for Cloud Applications
Mohan Baruwal Chhetri, Abdur Rahim Mohammad Forkan, Quoc Bao Vo, Surya Nepal, Ryszard Kowalczyk · 2019
Low-cost transient resources can be leveraged to reduce the cost of running applications on the cloud. However, they are susceptible to unilateral revocations by the provider, making them a risky choice for applications with strict performance requirements. Effective use of transient resources to minimize cloud costs without impacting the application performance remains an ongoing challenge for cloud consumers. In this paper, we address this challenge by proposing an approach for risk-aware, cost-optimal cloud resource allocation based on the concept of diversification. Contract diversification mitigates the risk to application performance due to resource revocation by procuring the required capacity as a mix of transient and non-transient resources. Resource diversification mitigates the risk further, by procuring the transient capacity as a heterogeneous mix of instance types. As supported by experimental results, the proposed approach achieves meaningful cost savings and risk mitigation.