Towards a Fault-Resilient Cloud Management Stack
Xiaoen Ju, Livio Soares, Kang Geun Shin, Kyung Dong Ryu · 2013
Cloud management stacks have become a new important layer in cloud computing infrastructure, simplifying the configuration and management of cloud computing en-vironments. As the resource manager and controller of an entire cloud, a cloud management stack has signifi-cant impact on the fault-resilience of a cloud platform. However, our preliminary study on the fault-resilience of OpenStack—an open source state-of-the-art cloud man-agement stack—shows that such an emerging software stack needs to be better designed and tested in order to serve as a building block for fault-resilient cloud envi-ronments. We discuss the issues identified by our fault-injection tool and make suggestions on how to strengthen cloud management stacks. 1