Amazon CloudWatch
Amazon CloudWatch is the monitoring service from AWS. It collects metrics, logs and events, displays them in dashboards and triggers alarms if thresholds are exceeded.
What is Amazon CloudWatch?
Amazon CloudWatch is AWS’s central monitoring and observability service. It answers the crucial question for ongoing operations: is my infrastructure currently working as it should? To do this, CloudWatch continuously collects metrics, logs and events from almost all AWS services and makes them analysable.
Without monitoring, you’re running an application blind: problems only become apparent when users complain. CloudWatch ensures that anomalies are spotted early – ideally before they turn into a disruption.
The building blocks of CloudWatch
- Metrics: Measurements such as CPU utilisation, memory usage or the number of requests over time.
- Logs: Centralised collection of log output from applications and services.
- Alerts: Notifications triggered when a metric exceeds or falls below a threshold.
- Dashboards: Customisable overviews that display the most important metrics at a glance.
From metrics to alarms
The real value of CloudWatch comes from alarms. An alarm monitors a metric and reacts automatically as soon as it leaves a defined range – for example, if CPU utilisation remains very high for a prolonged period or the error rate rises. An alarm can notify the team, but it can also trigger an automatic response, such as adding more servers via Auto Scaling.
Analysing logs in a targeted manner
CloudWatch not only collects logs, but also makes them searchable. With CloudWatch Logs Insights, large volumes of logs can be analysed using queries – for example, to find the cause of an error or identify unusual patterns. To ensure storage costs do not rise unnoticed, retention periods for logs should be set deliberately.
CloudWatch for SMEs
Monitoring is not an optional extra, but a prerequisite for reliable operations. CloudWatch is the obvious starting point because it works directly with all AWS services without the need for additional software. It is important to focus on a few meaningful alerts: a flood of notifications can quickly lead to even important warnings being overlooked. For deeper observability across multiple systems, CloudWatch can be combined with tools such as Prometheus and Grafana.
Frequently asked questions about Amazon CloudWatch
CloudWatch calculates metrics, alarms, logs and dashboards fed into the system individually. A permanent free quota covers basic monitoring. For medium-sized setups, the costs are usually in the single-digit to low double-digit euro range per month - with logs being the biggest cost factor.
Metrics are numerical measured values over time, such as CPU utilisation. Logs are the textual log outputs of applications and services. Metrics show that something is wrong, logs help to find out why.
Concentrate on a few, meaningful alarms that signal a real need for action. Too many alarms or alarms that are set too sensitively will lead to notifications being ignored - and even the important ones will go unnoticed.
CloudWatch is a solid starting point for many medium-sized setups. For distributed systems with many services, teams often supplement it with specialised tools such as Prometheus, Grafana or distributed tracing to make connections visible across system boundaries.
Related terms
Related services
Last updated: May 2026