Defining Operability for Web Services: Principles, Metrics, and Practices

This text addresses the important of production ready operability for web services specifically for the ones running in cloud. It outlines key system metrics and Java Virtual Machine (JVM) metrics that are essential for monitoring health and performance of cloud services. It also outlines key upstream and downstream metrics that are needed to ensure end to end monitoring is built before a service is said to be production ready. In addition to the metrics, we highlight the importance of other aspects such as centralized logging, distributed tracing that allows software engineers and other personas to quickly debug incidents and perform effective root cause analysis. We outline last mile aspects of what an effective incident response run book should contain which can be deemed effective ad ready to be used by on call engineer’s applications.