Datadog recommends the Serverless Framework Plugin for developers using the Serverless Framework to deploy their serverless applications. The plugin automatically enables instrumentation for applications to collect metrics, traces, and logs by:

• Installing the Datadog Lambda library to your Lambda functions as a Lambda layer.
• Installing the Datadog Lambda Extension to your Lambda functions as a Lambda layer (addExtension) or subscribing the Datadog Forwarder to your Lambda functions' log groups (forwarderArn).
• Making the required configuration changes, such as adding environment variables or additional tracing layers, to your Lambda functions.

Getting started

To quickly get started, follow the installation instructions for Python, Node.js, Ruby, Java, Go, or .NET and view your function's enhanced metrics, traces, and logs in Datadog.

After installation is complete, configure the advanced options to suit your monitoring needs.

Upgrade

Each version of the plugin is published with a specific set of versions of the Datadog Lambda layers. To pick up new features and bug fixes provided by the latest versions of Datadog Lambda layers, upgrade the serverless framework plugin. Test the new version before applying it on your production applications.

Configuration parameters

To further configure your plugin, use the following custom parameters in your serverless.yml:

Parameter	Description
site	Set which Datadog site to send data to, such as datadoghq.com (default), datadoghq.eu, us3.datadoghq.com, us5.datadoghq.com, ap1.datadoghq.com, or ddog-gov.com. This parameter is required when collecting telemetry using the Datadog Lambda Extension.
apiKey	Datadog API key. This parameter is required when collecting telemetry using the Datadog Lambda Extension. Alternatively, you can also set the DATADOG_API_KEY environment variable in your deployment environment.
appKey	Datadog app key. Only needed when the monitors field is defined. Alternatively, you can also set the DATADOG_APP_KEY environment variable in your deployment environment.
apiKeySecretArn	An alternative to using the apiKey field. The ARN of the secret that is storing the Datadog API key in AWS Secrets Manager. Remember to add the secretsmanager:GetSecretValue permission to the Lambda execution role.
apiKMSKey	An alternative to using the apiKey field. Datadog API key encrypted using KMS. Remember to add the kms:Decrypt permission to the Lambda execution role.
env	When set along with addExtension, a DD_ENV environment variable is added to all Lambda functions with the provided value. Otherwise, an env tag is added to all Lambda functions with the provided value. Defaults to the stage value of the serverless deployment.
service	When set along with addExtension, a DD_SERVICE environment variable is added to all Lambda functions with the provided value. Otherwise, a service tag is added to all Lambda functions with the provided value. Defaults to the service value of the serverless project.
version	When set along with addExtension, a DD_VERSION environment variable is added to all Lambda functions with the provided value. When set along with forwarderArn, a version tag is added to all Lambda functions with the provided value.
tags	A comma separated list of key:value pairs as a single string. When set along with extensionLayerVersion, a DD_TAGS environment variable is added to all Lambda functions with the provided value. When set along with forwarderArn, the plugin parses the string and sets each key:value pair as a tag on all Lambda functions.
enableXrayTracing	Set true to enable X-Ray tracing on the Lambda functions and API Gateway integrations. Defaults to false.
enableDDTracing	Enable Datadog tracing on the Lambda function. Defaults to true.
enableASM	Enable Datadog Application Security Management (ASM) on the Lambda function. Requires the Datadog extension to be present (using addExtension or manually added) and enableDDTracing. Defaults to false.
enableDDLogs	Enable Datadog log collection using the Lambda Extension. Defaults to true. Note: This setting has no effect on logs sent by the Datadog Forwarder.
monitors	When defined, the Datadog plugin configures monitors for the deployed function. Requires setting DATADOG_API_KEY and DATADOG_APP_KEY in your environment. To learn how to define monitors, see To Enable and Configure a Recommended Serverless Monitor.
captureLambdaPayload	Captures incoming and outgoing AWS Lambda payloads in the Datadog APM spans for Lambda invocations. Defaults to false.
enableSourceCodeIntegration	Enable Datadog Source Code Integration for the function. Defaults to true.
uploadGitMetadata	Enable Git metadata uploading for the function, as a part of source code integration. Set this to false if you have the Datadog Github Integration installed, as it renders Git metadata uploading unnecessary. Defaults to true.
subscribeToAccessLogs	Enable automatic subscription of the Datadog Forwarder to API Gateway access log groups. Requires setting forwarderArn. Defaults to true.
subscribeToExecutionLogs	Enable automatic subscription of the Datadog Forwarder to HTTP API and Websocket log groups. Requires setting forwarderArn. Defaults to true.
forwarderArn	The ARN of the Datadog Forwarder to be subscribed to the Lambda or API Gateway log groups.
addLayers	Whether to install the Datadog Lambda library as a layer. Defaults to true. Set to false when you plan to package the Datadog Lambda library to your function's deployment package on your own so that you can install a specific version of the Datadog Lambda library (Python or Node.js).
addExtension	Whether to install the Datadog Lambda Extension as a layer. Defaults to true. When enabled, it's required to set the apiKey and site.
exclude	When set, this plugin ignores all specified functions. Use this parameter if you have any functions that should not include Datadog functionality. Defaults to [].
enabled	When set to false, the Datadog plugin stays inactive. Defaults to true. You can control this option using an environment variable. For example, use enabled: ${strToBool(${env:DD_PLUGIN_ENABLED, true})} to activate/deactivate the plugin during deployment. Alternatively, you can also use the value passed in through --stage to control this option—see example.
customHandler	When set, the specified handler is set as the handler for all the functions.
failOnError	When set, this plugin throws an error if any custom Datadog monitors fail to create or update. This occurs after deploy, but will cause the result of serverless deploy to return a nonzero exit code (to fail user CI). Defaults to false.
logLevel	The log level for Datadog's lambda extension and tracing libraries. Set to DEBUG for extended logging. It does not affect which user logs are forwarded to Datadog.
skipCloudformationOutputs	Set to true if you want to skip adding Datadog Cloudformation Outputs for your stack. This is useful if you are running into the 200 output limit which can cause the stack creation to fail.
enableColdStartTracing	Set to false to disable Cold Start Tracing. Used in NodeJS and Python. Defaults to true.
coldStartTraceMinDuration	Sets the minimum duration (in milliseconds) for a module load event to be traced via Cold Start Tracing. Number. Defaults to 3.
coldStartTraceSkipLibs	optionally skip creating Cold Start Spans for a comma-separated list of libraries. Useful to limit depth or skip known libraries. Default depends on runtime.
subdomain	Set an optional subdomain to use for app URLs which are printed to output. Defaults to app.
enableProfiling	Enable the Datadog Continuous Profiler with true. Supported in Beta for NodeJS and Python. Defaults to false.
encodeAuthorizerContext	When set to true for Lambda authorizers, the tracing context will be encoded into the response for propagation. Supported for NodeJS and Python. Defaults to true.
decodeAuthorizerContext	When set to true for Lambdas that are authorized via Lambda authorizers, it will parse and use the encoded tracing context (if found). Supported for NodeJS and Python. Defaults to true.
apmFlushDeadline	Used to determine when to submit spans before a timeout occurs, in milliseconds. When the remaining time in an AWS Lambda invocation is less than the value set, the tracer attempts to submit the current active spans and all finished spans. Supported for NodeJS and Python. Defaults to 100 milliseconds.
enableStepFunctionsTracing	Enable automatic subscription of the Datadog Forwarder to Step Function log groups and Step Functions tracing. If no Step Function log groups are configured, then they are automatically created. Requires setting forwarderArn. Defaults to false.
propagateUpstreamTrace	When set to true, downstream Stepfunction invocation traces merge with upstream Stepfunction invocations. Defaults to false.
redirectHandlers	Optionally disable handler redirection if set to false. This should only be set to false when APM is fully disabled. Defaults to true.
To use any of these parameters, add a custom > datadog section to your serverless.yml similar to this example:

1custom:
2  datadog:
3    apiKeySecretArn: "{Datadog_API_Key_Secret_ARN}"
4    enableXrayTracing: false
5    enableDDTracing: true
6    enableDDLogs: true
7    subscribeToAccessLogs: true
8    forwarderArn: arn:aws:lambda:us-east-1:000000000000:function:datadog-forwarder
9    exclude:
10      - dd-excluded-function

Webpack

If you are using a bundler, such as webpack, see Serverless Tracing and Webpack.

TypeScript

You may encounter the error of missing type definitions. To resolve the error, add datadog-lambda-js and dd-trace to the devDependencies list of your project's package.json.

If you are using serverless-typescript, make sure that serverless-datadog is above the serverless-typescript entry in your serverless.yml. The plugin will automatically detect .ts files.

1plugins:
2  - serverless-plugin-datadog
3  - serverless-typescript

Disable Plugin for Particular Environment

If you'd like to turn off the plugin based on the environment (passed via --stage), you can use something similar to the example below.

1provider:
2  stage: ${self:opt.stage, 'dev'}
3
4custom:
5  staged: ${self:custom.stageVars.${self:provider.stage}, {}}
6
7  stageVars:
8    dev:
9      dd_enabled: false
10
11  datadog:
12    enabled: ${self:custom.staged.dd_enabled, true}

Serverless Monitors

There are seven recommended monitors with default values pre-configured.

Monitor	Metrics	Threshold	Serverless Monitor ID
High Error Rate	aws.lambda.errors/aws.lambda.invocations	>= 10%	high_error_rate
Timeout	aws.lambda.duration.max/aws.lambda.timeout	>= 1	timeout
Out of Memory	aws.lambda.enhanced.out_of_memory	> 0	out_of_memory
High Iterator Age	aws.lambda.iterator_age.maximum	>= 24 hrs	high_iterator_age
High Cold Start Rate	aws.lambda.enhanced.invocations(cold_start:true)/ aws.lambda.enhanced.invocations	>= 20%	high_cold_start_rate
High Throttles	aws.lambda.throttles/aws.lambda.invocations	>= 20%	high_throttles
Increased Cost	aws.lambda.enhanced.estimated_cost	↑20%	increased_cost

To Enable and Configure a Recommended Serverless Monitor

To create a recommended monitor, you must use its respective serverless monitor ID. Note that you must also set the DATADOG_API_KEY and DATADOG_APP_KEY in your environment.

If you’d like to further configure the parameters for a recommended monitor, you can directly define the parameter values below the serverless monitor ID. Parameters not specified under a recommended monitor will use the default recommended value. The query parameter for recommended monitors cannot be directly modified and will default to using the query valued as defined above; however, you may change the threshold value in query by re-defining it within the options parameter. To delete a monitor, remove the monitor from the serverless.yml template. For further documentation on how to define monitor parameters, see the Datadog Monitors API.

Monitor creation occurs after the function is deployed. In the event that a monitor is unsuccessfully created, the function will still be successfully deployed.

To create a recommended monitor with the default values

Define the appropriate serverless monitor ID without specifying any parameter values

1custom:
2  datadog:
3    addLayers: true
4    monitors:
5      - high_error_rate:

To configure a recommended monitor

1custom:
2  datadog:
3    addLayers: true
4    monitors:
5      - high_error_rate:
6          name: "High Error Rate with Modified Warning Threshold"
7          message: "More than 10% of the function’s invocations were errors in the selected time range. Notify @data.dog@datadoghq.com @slack-serverless-monitors"
8          tags: ["modified_error_rate", "serverless", "error_rate"]
9          require_full_window: true
10          priority: 2
11          options:
12            include_tags: true
13            notify_audit: true
14            thresholds:
15              warning: 0.05
16              critical: 0.1

To delete a monitor

Removing the serverless monitor ID and its parameters will delete the monitor.

To Enable and Configure a Custom Monitor

To define a custom monitor, you must define a unique serverless monitor ID string in addition to passing in the API key and Application key, DATADOG_API_KEY and DATADOG_APP_KEY, in your environment. The query parameter is required but every other parameter is optional. Define a unique serverless monitor ID string and specify the necessary parameters below. For further documentation on monitor parameters, see the Datadog Monitors API.

1custom:
2  datadog:
3    addLayers: true
4    monitors:
5      - custom_monitor_id:
6          name: "Custom Monitor"
7          query: "max(next_1w):forecast(avg:system.load.1{*}, 'linear', 1, interval='60m', history='1w', model='default') >= 3"
8          message: "Custom message for custom monitor. Notify @data.dog@datadoghq.com @slack-serverless-monitors"
9          tags: ["custom_monitor", "serverless"]
10          priority: 3
11          options:
12            enable_logs_sample: true
13            require_full_window: true
14            include_tags: false
15            notify_audit: true
16            notify_no_data: false
17            thresholds:
18              warning: 2
19              critical: 3

Breaking Changes

v5.0.0

• When used in conjunction with the Datadog Extension, this plugin sets service and env tags through environment variables instead of Lambda resource tags.
• The enableTags parameter was replaced by the new service, env parameters.

v4.0.0

• The Datadog Lambda Extension is now the default mechanism for transmitting telemetry to Datadog.

Working with serverless-plugin-warmup

This library is compatible at best effort with serverless-plugin-warmup. If you want to exclude the warmer function from Datadog, use the exclude feature of this library.

To properly package your application, this plugin must be listed after serverless-plugin-warmup in your serverless.yml file:

1plugins:
2  - serverless-plugin-warmup
3  - serverless-plugin-datadog

Opening Issues

If you encounter a bug with this package, let us know by filing an issue! Before opening a new issue, please search the existing issues to avoid duplicates.

When opening an issue, include your Serverless Framework version, Python/Node.js version, and stack trace if available. Also, please include the steps to reproduce when appropriate.

You can also open an issue for a feature request.

Contributing

If you find an issue with this package and have a fix, open a pull request following the procedures.

Community

For product feedback and questions, join the #serverless channel in the Datadog community on Slack.

License

Unless explicitly stated otherwise, all files in this repository are licensed under the Apache License Version 2.0.

This product includes software developed at Datadog (https://www.datadoghq.com/). Copyright 2021 Datadog, Inc.