AWS Lambda

Supported Library Modules

Script

Script

Prerequisite

To enable tracing for an application based on the Script module, sf-elastic-apm and sf-apm-lib must be available in your environment. These libraries can be installed by the following methods:

Add the below-mentioned entries in the requirements.txt file.

sf-apm-lib==0.1.1 
sf-elastic-apm==6.7.2 

For more information on AWS Lambda refer to: https://docs.aws.amazon.com/lambda/latest/dg/python-package-create.html#python-package-create-with-dependency

Configuration

Do the following steps to enable tracing for the AWS Lambda function.

1. Import the below-mentioned libraries.

   import elasticapm 
   from sf_apm_lib.snappyflow import Snappyflow 

2. Add the below code to get SnappyFlow Trace config outside the lambda handler method.

   sf = Snappyflow()
   SF_PROJECT_NAME = os.environ['SF_PROJECT_NAME'] 
   SF_APP_NAME = os.environ['SF_APP_NAME'] 
   SF_PROFILE_KEY = os.environ['SF_PROFILE_KEY'] 
   sf.init(SF_PROFILE_KEY, SF_PROJECT_NAME, SF_APP_NAME) 
   trace_config = snappyflow.get_trace_config() 

3. Add the custom instrumentation in the lambda handler function.

   def lambda_handler(event, context): 
      client = elasticapm.Client(service_name="<SERVICE_NAME_CHANGEME>", 
         server_url=trace_config['SFTRACE_SERVER_URL'], 
         verify_cert=trace_config['SFTRACE_VERIFY_SERVER_CERT'], 
         global_labels=trace_config['SFTRACE_GLOBAL_LABELS'] 
         ) 
      elasticapm.instrument()  
      client.begin_transaction(transaction_type="script") 
      # DO SOME WORK. No return statements. 
      client.end_transaction(name=__name__, result="success") 
      # RETURN STATEMENT e.g. return response 

4. To deploy the Lambda function, follow the README content in the code reference link.

   Sample code for reference: https://github.com/upendrasahu/aws-lambda-python-tracing-sample

5. Configure the Lambda function before trigger/invoke.

   • Add the environment variable SF_PROFILE_KEY and give the profile key copied from SnappyFlow as the value to the key.

   • Add the following environment variables: SF_APP_NAME and SF_PROJECT_NAME with appropriate values.