Using CloudWatch Metrics to Monitor API Gateway

Using CloudWatch Metrics to Monitor API Gateway

Use CloudWatch Metrics to monitor API Gateway directly from the AWS Console interface.

Objectives:

  • Monitor the number of requests, 4xx and 5xx errors, latency, integration latency, etc., of API Gateway
  • No need to create additional CloudWatch Log Groups or write manual code

Steps to follow:

  1. Access CloudWatch
  • Log in to the AWS Console
  • Type “CloudWatch” in the search bar → select CloudWatch
  1. Choose Metrics > All Metrics

metrics

  • In the left menu, choose Metrics → click All metrics
  • In the Browse section, select Lambda:

metrics

  1. Select the Metrics to Monitor:
  • Choose By Resource

metrics

  • Then select the function name(s) you want to monitor (we will focus on 3 main components: Throttles, Errors, and Duration)

metrics

  • Observe them on the chart

metrics

  • The 3 displayed metrics are:

    • Duration: The average execution time of the Lambda function (in ms) at each point in time. Displayed in green.
    • Errors: The number of errors that occurred at that point. Displayed in orange.
    • Throttles: The number of times the Lambda was throttled (exceeded allowed requests). Displayed in light blue.

  • When you hover over the data point at 07-30 09:35 UTC, CloudWatch shows:

    • Duration: 121.79 ms → Lambda took about 122 ms to execute

    • Errors: 0.4 → An average error rate (e.g., 2 errors in 5 requests or a rounded value)

    • Throttles: 0 → No requests were throttled at that time