Measuring to improve
Performance optimization, threat detection and mitigation, and ensuring reliable operation of any system, including edge solutions, is the ultimate goal of monitoring and observability. All the collection of metrics and logging and setting up alerts will not amount to much if the gathered information is not used to do root cause analysis (RCA) and fix what is wrong or improve a system’s performance. In that context, dashboards play an important role in the observability domain.
Dashboards should help with visualizing curated data, providing context, and offering a holistic view over time rather than at just a point in time. By displaying graphs and trend lines along with the actual metrics, dashboards help humans see the history of the data and the impact of an alarm when it occurred, and, most importantly, they add context and reason. Note that alarms should only be set for things that are important and they should be actionable.
