# AWS Lambda Instrumentation This package contains libraries to help instrument AWS lambda functions in your code. ## Using wrappers To use the instrumentation, configure `OTEL_INSTRUMENTATION_AWS_LAMBDA_HANDLER` env property to your lambda handler method in following format `package.ClassName::methodName` and use one of wrappers as your lambda `Handler`. In order to configure a span flush timeout (default is set to 1 second), please configure `OTEL_INSTRUMENTATION_AWS_LAMBDA_FLUSH_TIMEOUT` env property. The value is in seconds. Available wrappers: - `io.opentelemetry.instrumentation.awslambdacore.v1_0.TracingRequestStreamWrapper` - for wrapping streaming handlers (implementing `RequestStreamHandler`), enabling HTTP context propagation for HTTP requests When using known Lambda event types as parameters, use [aws-lambda-events-2.2](../../aws-lambda-events-2.2/library). ## Using handlers To use the instrumentation, replace your function classes that implement `RequestHandler` (or `RequestStreamHandler`) with those that extend `TracingRequestHandler` (or `TracingRequestStreamHandler`). You will need to change the method name to `doHandleRequest` and pass an initialized `OpenTelemetrySdk` to the base class. ```java public class MyRequestHandler extends TracingRequestHandler { private static final OpenTelemetrySdk SDK = OpenTelemetrySdk.builder() .addSpanProcessor(spanProcessor) .buildAndRegisterGlobal(); public MyRequestHandler() { super(SDK); } // Note the method is named doHandleRequest instead of handleRequest. @Override protected String doHandleRequest(String input, Context context) { if (input.equals("hello")) { return "world"; } return "goodbye"; } } ``` A `SERVER` span will be created with the name you specify for the function when deploying it. In addition, it is recommended to set up X-Ray trace propagation to be able to link to tracing information provided by Lambda itself. To do so, add a dependency on `io.opentelemetry.contrib:opentelemetry-aws-xray-propagator`. Replace `OPENTELEMETRY_VERSION` with the [latest release](https://search.maven.org/search?q=g:io.opentelemetry.contrib%20AND%20a:opentelemetry-aws-xray-propagator). Gradle: ```kotlin dependencies { implementation("io.opentelemetry.contrib:opentelemetry-aws-xray-propagator:OPENTELEMETRY_VERSION") } ``` Maven: ```xml io.opentelemetry.contrib opentelemetry-aws-xray-propagator OPENTELEMETRY_VERSION ``` ## Trace propagation Context propagation for this instrumentation can be done either with X-Ray propagation or regular HTTP propagation. If X-Ray is enabled for instrumented lambda, it will be preferred. If X-Ray is disabled, HTTP propagation will be tried (that is HTTP headers will be read to check for a valid trace context). ### X-Ray propagation This instrumentation supports propagating traces using the `X-Amzn-Trace-Id` format for both normal requests and SQS requests. X-Ray propagation is always enabled, there is no need to configure it explicitly. ### HTTP headers based propagation For API Gateway (HTTP) requests instrumented by using one of following methods: - extending `TracingRequestStreamHandler` or `TracingRequestHandler` - wrapping with `TracingRequestStreamWrapper` or `TracingRequestApiGatewayWrapper` traces can be propagated with supported HTTP headers (see ). In order to enable requested propagation for a handler, configure it on the SDK you build. ```java static { OpenTelemetrySdk.builder() ... .setPropagators(ContextPropagators.create(B3Propagator.injectingSingleHeader())) .buildAndRegisterGlobal(); } ``` If using the wrappers, set the `OTEL_PROPAGATORS` environment variable as described [here](https://github.com/open-telemetry/opentelemetry-java/blob/main/sdk-extensions/autoconfigure/README.md#propagator).