Essential Observability – Metrics, Events, Logs, and Traces (MELT)

If you cannot find any problems, then none exist! Are you sure about that? You probably wish this was true. However, in the systems visibility context, we can assume if we don’t find any problems, most likely, we have a blind spot. To illustrate that, how many times have an operations team heard about service degradation from its user first?

Observability, in making systems observable, is a notable feature of site reliability engineering but also one of its guiding principles. We can define it as the means of understanding the inner states of a system (or solution) by inspecting its outputs or signals. Although observability is an evolved telemetry model, where we want to collect measurements from a distance to minimally affect the measured system, it may require logging and tracing components to sharpen the system visibility. Observability has many characteristics, techniques, and considerations that...