Abstract
Virtualised environments such as cloud and edge computing architectures allow software to be deployed and managed through third-party provided services. Here virtualised resources available can be adjusted, even dynamically to changing needs. However, the problem is often the boundary between the service provider and the service consumer. Often there is no direct access to execution parameters at resource level on the provider’s side. Generally, only some quality factors can be directly observed while others remain hidden from the consumer. We propose an architecture for autonomous anomaly analysis for clustered cloud or edge resources. The key contribution is that the architecture determines possible causes of consumer-observed anomalies in an underlying provider-controlled infrastructure. We use Hidden Hierarchical Markov Models to map observed performance anomalies to hidden resources, and to identify the root causes of the observed anomalies in order to improve reliability. We apply the model to clustered hierarchically organised cloud computing resources. We illustrate use cases in the context of container technologies to show the utility of the proposed architecture.