CloudWatch is a centralized platform that allows customers to collect, monitor, and visualize their application logs, metrics, and X-Ray traces, all in a single dashboard. It is invaluable in providing a unified view of operational health for all the workloads the customer is managing.

There are three main areas of features that CloudWatch offers.

Infrastructure monitoring

Infrastructure monitoring covers the foundational aspects of metrics, alarms, and related insights.

CloudWatch metrics

Most AWS services emit free metrics by default that can be used to understand the performance of your systems and enable search and visualization capabilities. Customers can either use basic monitoring of metrics (free) or the detailed version (paid), which is only offered by some services. Depending on the time granularity of these metrics, the users could use them to understand trends and patterns of how certain parts of the system are working. Users can also leverage CloudWatch Metrics Insights, which is a SQL query engine that allows real-time analysis of metrics.

CloudWatch alarms

A CloudWatch alarm can be configured to monitor a certain metric or the result of a math expression that covers multiple metrics. For example, if an AWS service emits read IOPS and write IOPS independently, you can use metric addition to derive the total IOPS and configure an alarm whenever the total IOPS threshold is breached.

Alarms can be configured to send Amazon Simple Notification Service (SNS) notifications and start/ stop EC2 instances as a target action, for example.

Application monitoring

With application monitoring, CloudWatch offers additional insights into how your applications are working under the hood. Several services are offered in this category.

Synthetic monitoring

CloudWatch allows you to simulate an end user’s interaction with your application from multiple locations in the world – just as real users would use your application.

Service lens

Service lens aggregates all data from traces, metrics, logs, and alarms onto a single dashboard. This gives the user a unified view of not only the service map, which highlights the different components involved in the application workflow but also brings the bottlenecks and metrics for important data to the surface, such as requests per minute and average response times. This visibility allows you to easily pinpoint the needle in a haystack of data, and you can work back from this analysis to uncover the underlying problem.

CloudWatch real user monitoring

CloudWatch real user monitoring (RUM) can offer real user insights that give visibility into what the end user is experiencing. It’s a JavaScript code snippet that you can add to your application and start measuring key metrics immediately. The data that’s captured by the RUM agent and injected into your web page can extract page load times and client-side errors, among many other use cases.