A Resilience Test Case Design Method for Network Function Virtualization Platforms

摘要

With Cloud computing widely used in various applications in recent years, high availability(HA), has become a key characteristic for the quality of these cloud computing platforms. And there are already many fault-tolerant mechanisms that can keep services available when fault happens. However, whether the service performance is degraded under the occurrence of failures is not concerned. Such feature, is now called 'resilience', as the extension of high availability, and has become a hot topic in cloud computing. Based on former works of high availability testing on Network Function Virtualization(NFV) Platforms, this paper proposes a resilience test case design method. The method focuses on the service performance of NFV Platforms in face of high work-load and fault-load, and evaluates resilience of Virtual Network Function(VNF) services. Then a case study is conducted on an open-source NFV application, Clearwater. By comparing the test results to HA testing results, we learned more lessons on what failures impact the platform, which proved the feasibility and effectiveness of the proposed method.