𝞴 Lambda monitoring

Helios offers a wide range of deployment, observability & troubleshooting capabilities for AWS Lambda and other serverless functions.

• Deployment manually, through a Lambda layer or even a serverless plugin
• Observability & monitoring based on both the instrumented traces & metrics, as well as the raw logs from AWS
• Troubleshooting all errors - applicative or not - made simpler based on E2E traces and context propagated properly

Different installation options for Lambda observability

Install the Helios SDK in your services by leveraging any of the supported deployment options:

1. Manually setting environment variables in Lambda configuration (Go | Node.js | Python)
2. Using Helios' Lambda layers - the most updated version is available under Settings > General in the Helios app
3. Configuring Helios' serverless plugins

Lambda observability & monitoring

👉

AWS integration is required in order to get the complete Lambda observability & monitoring capabilities

Lambda overview

Each Lambda, represented as its own service in Helios, also provides a Lambda overview. It includes various trends & stats on metrics such as error trends, invocations, etc.

AWS Lambda status page

This Lambda status page provides a snapshot of what's going with the Lambdas used, across all regions. For each function and within the time frame selected it displays the last invocation, number of invocations, number of function errors, number of OOMs, number of timeouts, average duration, and also account ID, region and ARN. It is available under Cloud entities > Lambda.

To enable quicker troubleshooting and access to the full context of the invocations and errors - there are also links to the relevant traces in Helios so you can easily see E2E flows.

E2E trace visualization

Gaining intuitive visibility into an end-to-end flow with your Lambda functions is possible leveraging Helios' OpenTelemetry SDK and showing the data and appropriate context.

Lambda troubleshooting

Alerts

Users can customize and control exactly what data is important to them with each Lambda by leveraging the labels & alerts in Helios to save search queries and detect behaviors that are of interest. Customization can be done based on either applicative events, or Lambda metrics from AWS.

Automatic access to CloudWatch logs

For each span it's possible to retrieve the relevant logs from CloudWatch with a short click of a button.

Flow replay

Similar to other types of traces in Helios, each Lambda trace offers several quick troubleshooting actions. A key action is the ability to replay a specific flow by automatically generating the code and then being able to configure it an use it to reproduce an issue, run it in a different environment, investigate the root cause, and finally verify that it's working properly.