An Overview of Virtual Machine Monitoring Techniques and Observability

Abstract

The adoption of cloud computing systems is rapidly increasing, necessitating robust monitoring to ensure optimal performance and a comprehensive understanding of the internal state of these systems. Effective monitoring is crucial for early anomaly detection and preventing potential failures. Unlike grid computing, cluster computing, and high-performance computing. Cloud computing introduces unique features such as scalability and elasticity, which require specialised monitoring approaches. Existing literature indicates that traditional monitoring tools from previous computing paradigms have been adapted for cloud systems. However, these tools often fall short due to the distinct characteristics of cloud environments. This paper identifies a significant gap in research on monitoring techniques specifically designed for virtual resources and highlights the importance of observability in cloud systems. Observability, which facilitates root cause analysis, is essential for quickly resolving faults by examining the internal state of the cloud system. This paper explores the primary characteristics of cloud monitoring tools, providing examples of currently used tools and their features. Additionally, it surveys virtual machine monitoring techniques and emphasises the critical role of observability in ensuring the rapid resolution of issues through logs, metrics, traces, and dependencies.