Choosing the Right Observability Signal Source
Explore how to identify the ideal observability signal source to diagnose AWS incidents efficiently. Understand the roles of CloudWatch, Amazon Managed Prometheus, CloudTrail, and AWS Health Dashboard in confirming service impact, auditing changes, and detecting provider events. This lesson guides you in matching evidence sources to specific monitoring questions to improve incident response accuracy and reduce investigation time.
Deciding whether an incident traces back to a recent change, an AWS-side event, or genuine load on a workload depends entirely on picking the evidence source that can actually answer that specific question. Five different sources exist for exactly this reason, and each one answers a different question quickly while answering the others poorly or not at all.
CloudWatch confirms service and infrastructure impact through metrics, logs, and alarms. Amazon Managed Grafana confirms how multiple signals line up by rendering dashboards across sources, without creating any new data of its own. Amazon Managed Service for Prometheus confirms workload telemetry that CloudWatch doesn't have, by storing and querying Prometheus-style metrics. CloudTrail confirms whether a principal changed a resource. AWS Health Dashboard confirms whether AWS is reporting an event that can ...