Observability
What AWS Provides by Default
Without any configuration, EC2 publishes basic CloudWatch metrics at 5-minute intervals. These cover CPU utilization, network in/out, and disk read/write operations. They do not cover memory usage or disk space utilization — those require the CloudWatch Agent to be installed and configured on the instance.
To view the default metrics, navigate to EC2 > Instances > select your instance > Monitoring tab. You will see graphs for CPU utilization and network traffic.
For a production instance, 5-minute granularity is often insufficient for diagnosing short-lived performance spikes. Detailed monitoring (1-minute granularity) is available at an additional cost of approximately $3.50 per instance per month. For this lab, the default is sufficient.
System Log Access
If an instance fails to boot or you cannot connect, the first diagnostic tool is the EC2 system log. In the EC2 console, select your instance, then Actions > Monitor and troubleshoot > Get system log. This captures the serial console output from boot — kernel messages, cloud-init output, and any SSH daemon startup messages. If SSH is not starting or the instance is failing to initialize, the system log will contain the reason.
In this section, I confirmed:
0 of 3 completed