What is Cloudwatch on AWS

Amazon CloudWatch Features

Monitor

Uniform functional view with dashboards

With Amazon CloudWatch dashboards, you can create reusable charts and visualize your cloud resources and applications in a unified view. You can graph metrics and historical data side-by-side on a single dashboard to quickly get context and move from diagnosing the problem to understanding the root cause. For example, you can visualize key metrics like CPU and memory usage and compare them to capacity. You can also relate the logging pattern of a particular metric and set alarms to proactively raise if there are performance or functional issues. This gives you a system-wide insight into the functional status and enables you to resolve problems quickly and reduce the average time to resolution (mean time to resolution, MTTR).

Composite alarms

With Amazon CloudWatch's composite alarms, you can combine multiple alarms and reduce alarm noise. If an application problem affects multiple resources in the application, you will receive a single alert notification for the entire application instead of a notification for each affected service component or resource. This way, you can continue to focus on finding the root cause of the operational problems to reduce application downtime. You can create an overall health for a grouping of resources, such as For example, deploy an application, AWS Region, or Availability Zone.

High resolution alarms

Amazon CloudWatch alarms allow you to define a threshold for metrics and trigger an action. You can create high resolution alarms, set a percentile as a statistic and, if necessary, specify an action or ignore the alarm. For example, you can create alarms for Amazon EC2 metrics, set notifications, and take one or more actions to detect and shut down unused or underutilized instances. Real-time alerts for metrics and events enable you to minimize downtime and potential business impact.

Correlation between logs and metrics

Applications and infrastructure resources generate vast amounts of functional and monitoring data in the form of logs and metrics. With Amazon CloudWatch, you can not only access these data sets and visualize them on a single platform, but also relate metrics and logs. This helps you move quickly from diagnosing the problem to understanding the root cause. You can e.g. For example, you can relate a log pattern (such as an error) to a specific metric and set alarms that are actively issued in the event of performance or functional problems.

Application Insights for .NET and SQL Server applications

Amazon CloudWatch Application Insights for .NET and SQL Server enables you to easily monitor .NET and SQL Server applications to get an overview of the health of such applications. It enables the discovery and establishment of key metrics and protocols for the application resources and technologies you use, such as: B. Database, web (IIS) and application server, operating system, load balancer, queues, etc. This telemetry data is constantly monitored in order to detect and correlate anomalies and errors and to inform you of problems in the application. To make troubleshooting easier, it creates automated dashboards for detected issues that break down correlated metric anomalies and log errors, as well as additional insights that suggest possible causes. This allows you to quickly take countermeasures and ensure that your applications are working properly and that end users are not affected.

Insights into container monitoring

In Container Insights, you can access automated dashboards in the CloudWatch console. These dashboards summarize compute performance, errors and alarms by cluster, pod / task and service. Dashboards for nodes / EC2 instances and namespaces are also available for Amazon EKS and k8s. Each dashboard summarizes the list of running pods / tasks or containers by CPU and memory for the selected time window. Depending on the time slot and the selected pods / tasks or containers, you will get a deeper insight into the application logs, AWS X-Ray traces and the performance events.

Insights into lambda monitoring

CloudWatch Lambda Insights provides automated dashboards in the CloudWatch console. These dashboards summarize the computing power and the errors. Each dashboard contains the list of metrics for the selected time slot and allows you to delve deeper into application logs, AWS X-Ray traces, and performance events in context based on the time slot and function selected.

Anomaly detection

Amazon CloudWatch Anomaly Detection uses machine learning algorithms to continuously analyze the data of a metric and to detect behavioral anomalies. This allows you to create alarms that automatically adjust thresholds based on natural metric patterns, including time of day, seasonality by day of the week, or trend changes. Anomaly detection tapes can also be used to visualize metrics on dashboards. This enables you to monitor, isolate, and fix unexpected deviations in your metrics.

ServiceLens

With Amazon CloudWatch ServiceLens you can visualize and analyze the health, performance and availability of your applications. CloudWatch ServiceLens brings together CloudWatch metrics and protocols as well as traces from AWS X-Ray into a central view of your applications and their dependencies. This enables you to quickly map performance bottlenecks, isolate the causes of application problems, and identify affected users. CloudWatch ServiceLens gives you insights into three main areas of your applications: infrastructure monitoring (metrics and logs provide information about the resources that support your applications), transaction monitoring (traces provide information about dependencies between resources), and end-user monitoring (canaries monitor the endpoints and notify you when the end-user experience worsens). CloudWatch ServiceLens offers a service overview in which the links of all your resources are visualized in their context. Thoroughly analyze correlated monitoring data through an intuitive user interface.

Synthetics

Amazon CloudWatch Synthetics makes it easier to monitor application endpoints. The feature runs every minute tests on your endpoints around the clock and notifies you if the application endpoints are not behaving as expected. With the customizable tests you can e.g. Check for availability, latency, transactions, broken links, step-by-step task completion, page loading errors, loading latencies for UI resources, complex assistant processes and checkout processes in your applications. You can also use CloudWatch Synthetics to isolate problematic application endpoints and map them to underlying infrastructure problems to reduce mean troubleshooting time. With this new feature, CloudWatch can now capture canary traffic. This allows you to continuously verify the customer experience, even when there is no customer traffic for your applications, and identify any problems before they affect customers. CloudWatch Synthetics supports monitoring of REST APIs, URLs and website content and checks for unauthorized changes through phishing, code injection and cross-site scripting.

Stream metrics

With Amazon CloudWatch Metric Streams, you can create continuous, near real-time streams of metrics toward a destination of your choice. Metrics Streams makes it easier to send CloudWatch metrics to external service providers through an Amazon Kinesis Data Firehose HTTP endpoint. You can create an ongoing, scalable stream, including the latest CloudWatch metrics, to power dashboards, alarms, and other tools that rely on accurate and timely metric data. You can easily direct your metrics to your data lake on AWS, like on Amazon Simple Storage Service (S3), and you can immediately start analyzing usage or performance with tools like Amazon Athena.