Skip to main content

Asynchronous Python Executor

Our Run Asynchronous Python Code (Advanced) module allows you to run Python code for up to 30 minutes (compared to only 3 minutes for the normal Python executor). The tradeoff is that you need to get the result asynchronously via a webhook or polling. You can find the API docs here.

Overview​

Async Python Module in Make

You can configure three things here:

  1. Code is your Python code itself. This will be executed by our code executor, just like when you run python yourself locally.
warning

In contrast to the non-async Python executor, you can use print in the async Python executor. Assigning data to the global result variable will have no effect.

  1. Requirements are the PyPI packages you want to use. These are passed verbatim to pip (which you can also use locally).
warning

Don't pass standard library packages here, like json. These are already installed (like in every Python environment), and won't be found on PyPI. Remember, the Requirements are passed directly to pip.

  1. Send To is your webhook URL. See below for more information.

Getting Results​

Unfortunately, platforms like Make impose HTTP timeouts, and we therefore cannot execute code in a single connection for arbitrary amounts of time. Our solution is to give you the results asynchronously. This does complicate your workflow a little, but there is currently no other option.

This section will describe the two different options you have to retrieve your results after your code has finished executing.

Webhooks​

If you provide a webhook URL in the Send To field, the async Python executor will send an HTTP POST request to this webhook once your Python code has finished executing. The request body will look like this:

async_python_webhook_body.json
{
  "status": "success",
  "result": "Hello, world!\n",
  "endpoint": "/code/async-python"
}

You can create a webhook in Make relatively easily using the Custom Webhook module:

Custom Webhook Module in Make

The shown webhook URL has to be pasted into the Send To field. The rest of your workflow can continue from this webhook node and use the results of your Python code.

tip

It makes sense to print a JSON value to the console and then parse that with the Make module Parse JSON to easily process the result.

Task Status​

warning

In most cases, it is significantly easier to use webhooks for this. Only use this module if you are sure that webhooks don't work for your purpose.

If you do not want to use a webhook, you can also poll the results regularly.

Task Status Module in Make

You can use the task ID provided to you by the Run Asynchronous Python Code (Advanced) module, and you will get the current status, and the results if your code is done already.

Pricing​

The asynchronous Python executor can run up to 30 minutes, and costs 50 credits per every started 3 minutes of execution. If you execute code for 2 minutes, this will cost 50 credits, 7 minutes will cost 150, and 30 minutes will cost 500.