Monitoring AWS Lambda Invocations

In this guide, you will:

• Use the AWS Console to monitor AWS Lambda invocations in self-hosted environments.

Monitoring AWS Lambda invocations is crucial for detecting performance bottlenecks, optimizing execution efficiency, and managing AWS costs effectively. Worker Lambda functions process events, and prolonged execution times can result from high concurrency, database overload, or excessive SQS message throughput.

When durations exceed defined thresholds, they increase costs and may indicate inefficiencies. Proactive monitoring helps ensure smooth execution, efficient resource utilization, and minimized costs.

Prerequisites

• Access to the Guardrails AWS account with ReadOnly privileges.
• Familiarity with AWS Console.

Step 1: Log in to AWS Console

Open the AWS Console and navigate to the Lambda service in the region where Guardrails is deployed.

Step 2: Navigate to Dashboards

From the left navigation menu, select Dashboards.

Step 3: View Lambda Invocations

In Account-level metrics, select Invocations and set the desired date range.

Step 4: Identify Invocation Spikes

Analyze the graph for spikes in invocation counts over the selected time range. In this example, a spike is observed starting from 01/29/2025.

Step 5: Investigate and Resolve Spike

A spike in Lambda invocations may be caused by increased concurrency, database overload, or excessive SQS message throughput, often due to misconfigured policies or external API events.

To identify the root cause, analyze event floods for unusual spikes in external events.

Common causes include:

• Unintended API calls
• Excessive policy evaluations
• High-frequency control triggers

Once identified, apply necessary fixes as recommended by Turbot support.

Step 6: Review

• Ensure that the Lambda invocation spike stabilizes and returns to normal levels.

Next Steps

Explore additional monitoring guides for Guardrails Enterprise:

• Learn how to Monitor Alarms.
• Learn how to Diagnose Control Errors.

Troubleshooting