phhelper 0.9.18

Python Handler Helper for Lambda

Overview

Simplify best practice for Lambda in Python, handling Lambda events and errors with detailed and standarized logs.

Features

• Dead simple to use, reduces the complexity of writing a Lambda with Python runtime

• Guarantees that Source will get a response even if an exception is raised

• Sends meaningful errors to Cloudwatch in the case of a failure

• Threading enables best runtime performance for events with multiple records

• JSON logging that includes request id’s, event id’s and source to assist in tracing logs relevant to a particular event

Installation

Install into the root folder of your lambda function

cd my-lambda-function/
pip install phhelper -t .

Example Usage

from phhelper import aws_lambda_helper
import json
import time

@aws_lambda_helper.handler
def lambda_handler(event, context):
    context.logging.info('info_test')
    time.sleep(0.5)
    context.logging.error('error_test')
    time.sleep(0.5)
    context.logging.debug('debug_test')

    return {
        'statusCode': 200,
        'body': json.dumps('Hello from Lambda!')
    }

Threading

If the event source send multiple records to be processed, you can enable Multithreading processing, by creating a Environment variable called THREADING_ENABLED with value TRUE. This will make a loop into event records, starting a thread for each record. Your handler will receive each record in a separeted call, inside a thread model.

Logging

You can define the general log verbosity level using a Environment variable called LOG_LEVEL and the boto3 log level using a Environment variable called BOTO_LOG_LEVEL.

The valid values for both Environment variables are: * DEBUG * INFO * WARNING * ERROR * CRITICAL

The default values are:

LOG_LEVEL = 'ERROR'
BOTO_LOG_LEVEL` = 'CRITICAL'

Credits

Decorator implementation inspired by https://github.com/aws-cloudformation/custom-resource-helper

Log implementation inspired by https://gitlab.com/hadrien/aws_lambda_logging

Multiprocessing implementation inspired by https://medium.com/@urban_institute/using-multiprocessing-to-make-python-code-faster-23ea5ef996ba

License

This library is licensed under the MIT License.