Monitoring AWS Lambda Performance with CloudWatch

Introduction

CloudWatch provides metrics and logging for AWS Lambda to monitor performance, errors, and invocations in real time.

Prerequisites

AWS CLI configured
Lambda functions deployed

Step 1: Enable Enhanced Metrics

In Lambda console, enable Active Tracing with X-Ray or choose Enhanced Monitoring.

Step 2: Configure CloudWatch Logs

Ensure function has IAM role with logs:CreateLogGroup and logs:CreateLogStream permissions.

Step 3: Create CloudWatch Dashboard

aws cloudwatch put-dashboard \
  --dashboard-name LambdaDashboard \
  --dashboard-body file://dashboard.json

dashboard.json example:

{
  "widgets": [
    {
      "type": "metric",
      "x": 0, "y": 0, "width": 24, "height": 6,
      "properties": {
        "metrics": [
          ["AWS/Lambda", "Invocations", "FunctionName", "myLambda"],
          [".", "Errors", ".", "."]
        ],
        "period": 60,
        "stat": "Sum",
        "view": "timeSeries",
        "region": "us-east-1"
      }
    }
  ]
}

Step 4: Set Alarms

aws cloudwatch put-metric-alarm \
  --alarm-name LambdaErrorsAlarm \
  --metric-name Errors \
  --namespace AWS/Lambda \
  --statistic Sum \
  --period 300 \
  --threshold 1 \
  --comparison-operator GreaterThanOrEqualToThreshold \
  --dimensions Name=FunctionName,Value=myLambda \
  --evaluation-periods 1 \
  --alarm-actions arn:aws:sns:us-east-1:123456789012:Alerts

Step 5: Analyze Logs

Use CloudWatch Logs Insights:

defaultView: SELECT @timestamp, @message
| filter @message LIKE "ERROR"
| sort @timestamp desc
| limit 20

Summary

CloudWatch metrics, dashboards, and alarms enable comprehensive monitoring of Lambda performance, ensuring reliability and quick issue detection.