This is an example of a simple, traced Lambda function using the SignalFx Lambda Wrapper and Jaeger Python tracer with SignalFx. See example.py for the example code.
A Python deployment package for AWS Lambda is a .zip file consisting of your application code and any dependencies. To prepare this project and create the deployable archive follow these instructions from this document's parent directory:
$ pwd # Confirm this is <your_local_path/tracing-examples/aws-lambda/jaeger-python>
$ pip install -r requirements.txt -t .
$ find . -name "*.py[co]" -delete
$ find . -name "*.dist-info" -exec rm -r "{}" \;
$ # If you are running your Lambda function in the Python 3.6 runtime environment
$ # you'll need to remove the concurrent package as its code is evaluated at load time
$ # and raises a SyntaxError for using the obsoleted raise Exception tuple syntax. This
$ # is part of the standard library for Python 3.2+ so is not necessary.
$ # rm -rf ./concurrent
$ zip -r my_traced_python_lambda.zip *
The resulting my_traced_python_lambda.zip can be uploaded to S3 or in your browser via the AWS Console
during function creation. Register your handler as example.request_handler and don't forget to set the
SIGNALFX_ENDPOINT_URL environment variable to point to your Gateway. Set
SIGNALFX_SERVICE_NAME to signalfx-lambda-python-example or something else
descriptive. You should be able test the application with the following test
event:
{
"queryStringParameters": {
"choice": "00"
}
}
The execution result should succeed with a response similar to:
{
"body": "{\"result\": \"You lost! The ball landed on 18.\", \"trace_id\": \"6527113c86f90ce9\", \"choice\": \"00\"}",
"statusCode": 404
}