chili-pepper

Asynchronous Serverless Task Execution

Chili-Pepper is a simple framework that makes it easy to execute tasks without interrupting the main flow of your application. It handles serverless deployment and task execution. It allows you to run important functions in parallel with your main application with zero downtime, zero maintenance and infinite scaling.

Command

usage: chili [-h] [--app APP] {deploy} ...

Serverless asynchronous tasks

positional arguments:
{deploy}           Chili-Pepper commands
    deploy           Deploy functions to serverless provider

optional arguments:
-h, --help         show this help message and exit
--app APP, -A APP  The Chili-Pepper application location

Getting Started

Installing Chili-Pepper

pip install chili-pepper

Serverless Provider

Next, you need to configure your serverless provider credentials.

Amazon Web Services (AWS)

You can provide AWS credentials in many ways, including environment variables, your aws credentials file, or a server role. See the boto3 documentation about credential configuration for details.

These credentials need to be allowed to create, execute and delete AWS Lambda functions and IAM roles.

You will also need an S3 bucket (with versioning enabled).

Azure

Microsoft Azure Cloud is not supported at this time, but there are plans to support it in the future.

Google Cloud

Google Cloud is not supported at this time, but there are plans to support it in the future.

Creating the Chili-Pepper App

The Chili-Pepper app will be used to identify and deploy your functions to the serverless cloud provider.

app = ChiliPepper(app_name="demo")

Of course, the app variable can have any name - we'll be calling it app in these examples.

AWS Configuration

You need to pass these required AWS specific configs to the app. For a full list of AWS configuration options, see Aws Configuration.

Bucket

app.conf["aws"]["bucket_name"] = "my-chili-pepper-bucket"

This bucket will be used for storing AWS Lambda deployment packages. You should enable versioning on the bucket.

Runtime

app.conf["aws"]["runtime"] = "python3.7"

AWS Lambda supports several python runtimes. See the lambda runtime documentation for a full list. You must pass the "Identifier" for the runtime of your choice to the Chili-Pepper app config.

Creating a Task

You can use the Chili-Pepper app to identify tasks that should be run.

@app.task()
def my_task(event, context):
    return f"Hello {event['name']}!"

Deploying

Before you can asynchronously call your task, you must deploy it to your cloud provider.

chili deploy --app my_module.tasks.app

AWS Deployment

Calling deploy will create a zipfile containing your code as well as any python dependencies. Chili-Pepper will then use upload that zipfile to your S3 bucket, and use Cloudformation to create an AWS Lambda function for each of the tasks you identified with the app.task() decorator.

Calling your task

Now that you've deployed your tasks to the cloud, you can call them asynchronously.

task_result = my_task.delay({"name": "Jalapeno"})
print(task_result.get())

This will print Hello Jalapeno!, after executing my_task in a serverless function.

Support

Chili-Pepper is built by a 1 person team supported by these awesome backers, supporters and sponsors. If you use Chili-Pepper, I would love to hear from you! And, if I have earned your support, please consider backing me on Patreon.

.. toctree::
   :maxdepth: 1
   :caption: Contents:

   config
   API Docs <modules>
   backers
   license

william-richard/chili-pepper