What is AWS monitoring?

With over 30% of the cloud market, AWS runs more infrastructure than any other cloud infrastructure provider.¹ Companies use AWS for web hosting, data storage, big data analytics, mobile app development and enterprise IT services.

For these organizations, monitoring their workloads on AWS is critical. It enables them to track performance metrics, system logs and events to help ensure that AWS environments perform as expected.

Without AWS monitoring, performance issues can multiply undetected, costs can spiral from overprovisioned resources and security vulnerabilities can remain exposed. For example, excessive network traffic might overload a central processing unit (CPU) and cause bottlenecks. Or misconfigured cloud storage containers might expose sensitive data through public access.

AWS monitoring tools can identify these issues and trigger automated responses—such as invoking AWS Lambda functions for remediation—or alert teams for manual troubleshooting. This helps organizations maintain optimal performance, reduce costs, strengthen their security posture and make data-driven decisions about their infrastructure.

AWS monitoring helps ensure the reliability, availability and performance of AWS cloud resources. It provides visibility into AWS infrastructures so organizations can proactively detect and resolve problems, reducing disruptions and minimizing downtime.

Monitoring helps reveal the operational health of AWS cloud-based resources, including compute capacity, storage systems and network infrastructure.

For example, if a server experiences heavy load from customer traffic, monitoring triggers automatic scaling to add more servers. This helps prevent application crashes and ensure consistent response times even during traffic spikes, so applications remain accessible to end users. 

By collecting metrics on AWS cloud resources, monitoring can help optimize configurations for increased speed and efficiency.

For example, if users in Asia experience slow page loads, monitoring tools can cache content in Singapore, closer to their location. This helps reduce latency and enables faster page loads, ensuring smoother streaming and improving the user experience.

Cloud costs are often one of the biggest line items on an organization’s IT budget.

Monitoring can identify underutilized or unnecessary AWS resources to help optimize cloud costs. For example, a virtual server using only 10% of its allocated memory for its workload can risk wasting 90% of memory costs. AWS monitoring can help automatically right size instances and shut down idle resources during off-peak hours. 

Monitoring detects suspicious activities or unusual changes to AWS infrastructures that might indicate security threats, such as unauthorized API calls, unusual data transfers and configuration changes.

For example, when monitoring tools detect repeatedly failed login attempts, they can block the source IP address and trigger security notifications.

Monitoring can identify potential problems before they occur and impact end users. This can include approaching capacity limits, degrading performance trends and expiring SSL certificates.

For example, monitoring a database approaching over 90% storage capacity triggers alerts to developers to add more space before database failure.

AWS monitoring continuously collects and analyzes data from AWS infrastructure.

Monitoring data typically includes performance metrics, resource utilization and error rates from servers, databases and applications, along with system logs such as API calls, configuration changes, network activity and security events.

Tools analyze this data to identify trends, detect anomalies and visualize performance in real time through dashboards.

AWS monitoring tools can then alert teams to potential issues for troubleshooting and root cause analysis. They can also automatically resolve some issues, including adding resources or restarting AWS services.

For example, when customers abandon online shopping carts due to failed checkouts, AWS monitoring can identify payment gateway errors and alert the DevOps team. They can then investigate and correct the outdated timeout setting, minimizing revenue loss.

AWS monitoring pricing typically scales with the number of custom metrics, log ingestion volumes and analysis frequency, depending on the services used.

To provide comprehensive coverage, AWS monitoring typically tracks metrics across four core areas:

• Infrastructure performance and usage
• Application performance and behavior
• Security and compliance
• Operational and business metrics

AWS monitoring automatically tracks performance and resource consumption across AWS infrastructure components.

These metrics help teams identify overprovisioned resources, predict capacity needs and detect performance degradation before it impacts users.

Amazon CloudWatch, the foundation of AWS monitoring, is the primary tool for gathering this data. It automatically collects, aggregates and analyzes metrics across all AWS services. Other AWS tools like X-Ray and CloudTrail integrate with CloudWatch to provide a unified view of system health, performance and security.

EC2 provides virtual servers that run applications in the cloud with on-demand computing capacity.

Key EC2  metrics include:

• CPU utilization: Percentage of allocated compute units being used
• Disk I/O: Rate of reads and writes from attached storage volumes
• Network traffic: Data flowing in and out of the EC2 instance
• Status checks: System and instance health checks that verify reachability and detect hardware or software issues 

ECS and EKS, Amazon’s native container services, deploy and orchestrate containerized applications at scale.

Key container metrics include:

• Resource utilization: CPU, memory usage and network usage
• Performance logs: Diagnostic information such as container restart failures for root-cause analysis
• Network performance: Network traffic between containers, services and external resources

RDS provides managed databases in AWS cloud environments.

Key RDS metrics include:

• Database connections: Number of active connections to each RDS database instance
• Query latency: Time to complete database queries
• Storage usage: Disk space consumed by databases

Lambda provides serverless computing for tasks such as resizing images or updating databases.

Key Lambda metrics include:

• Invocation count: Number of times a function executes
• Duration: Time to complete each function
• Errors: Number of function execution failures

ELB distributes incoming traffic to appropriate cloud resources to help maintain high availability.

Key ELB metrics include:

• Latency: Time requests take to process through the load balancer
• Request count: Total requests the load balancer handled  
• Active connection count: Concurrent connections from clients to targets
• Processed bytes: Total bytes processed by the load balancer

AWS performance monitoring can track application behavior but only with manual configuration and code instrumentation.

Applications rarely exist in isolation—they connect to external payment processors, third-party APIs and non-AWS databases.

Monitoring these interactions reveals whether performance problems stem from AWS resources or external dependencies. It also provides insight into how end users interact with applications.

AWS X-Ray, a distributed tracing service, is the primary tool for collecting these metrics. When code is instrumented with its SDK, X-Ray traces requests as they move through applications, providing visibility into latency, errors and bottlenecks across AWS services.

Key application performance monitoring metrics include:

• Request rates: Volume and patterns of incoming requests, revealing traffic spikes and usage trends

• Error rates: Failed transactions, crashed processes and broken API connections

• User monitoring: Real user experience data—page load times, JavaScript errors, click through rates

• Distributed tracing: End-to-end request journeys across microservices, pinpointing where bottlenecks or failures occur 

AWS monitoring automatically tracks activity and configuration changes within an AWS account. These metrics help identify who accessed sensitive data, detect policy violations and prove regulatory compliance.

Different AWS Security services provide different security benefits, such as audit trails, configuration tracking and threat detection. Some, like CloudTrail, log basic activity by default, while others—such as GuardDuty and Config—require explicit setup.

Key AWS-native security services include:

• AWS CloudTrail: Tracks every API call across AWS environments, logging who did what and when—essential for auditing and incident investigation.

• AWS Config: Records how AWS resources are configured and tracks changes over time, helping prove compliance and with policies.

• Amazon Inspector: A security and vulnerability management service that automatically scans EC2 instances and containers for known vulnerabilities and deviations from security best practices.

• Amazon GuardDuty: A continuous threat detection service that uses machine learning to detect malicious activity, including cryptocurrency mining, unauthorized access attempts and unusual API calls across an AWS account.

Key security metrics include:

• API calls: API calls are communications between AWS APIs and software programs to request data, services or actions. Source, destination and timing are recorded for all calls.

• Unauthorized access: Any unusual or malicious login activity.

• Configuration changes: Changes to AWS resource configurations for compliance and auditing. For example, changes to user permissions or data encryption might affect compliance with security policies or data protection regulations.

• Network traffic analysis: Network traffic across virtual private clouds (VPCs) to help spot security threats such as misconfigurations or data exfiltration attempts.

AWS monitoring can track business metrics but requires custom configuration, where organizations define and send them to AWS monitoring tools as custom metrics.

These metrics—such a revenue, order fulfillment time or customer satisfaction—can help connect technical performance to business outcomes, justify cloud spending and identify how system issues impact revenue.

Key operational and business metrics include:

• Customer service response time: How quickly systems respond to customer requests

• Order fulfillment time: Average time for e-commerce orders to be processed and shipped

• User activity: Daily and monthly active users, new sign-ups and session counts

• Revenue: Total monetary value of purchases over a specific period

• AWS billing: Anomaly detection can identify unusual usage spikes so teams can make cost-effective adjustments and meet budget thresholds

Monitoring helps answer what’s wrong, while observability helps explain why it’s happening.

AWS monitoring solutions typically include AWS observability capabilities. Both work together to solve problems and maintain reliability.

Monitoring captures predefined metrics such as CPU usage and network latency, offering a snapshot of system performance but little insight into why an issue occurs. For example, monitoring might identify high CPU utilization on a web server but not the root cause. 

Observability correlates multiple telemetry data types—metrics, events, logs and traces (MELT data)—to provide a real-time view of AWS environments. It evolved from traditional monitoring to handle the complexity of cloud-native architectures.