hello-world

Hello World - Render Workflows (Python)

This hello-world example demonstrates three foundational workflow patterns:

• A minimal task definition (calculate_square)
• A task that chains runs of another task (sum_squares)
• A task with custom retry behavior (flip_coin)

What You'll Learn

• How to define tasks with @app.task
• How to run one task from another with ctx.run
• How to run tasks in parallel with asyncio.gather
• How to customize retry behavior with Retry

Example Tasks

calculate_square(ctx, a: int) -> int

The smallest possible task: takes one integer and returns its square. Like every task, it receives a TaskContext as its first parameter.

sum_squares(ctx, a: int, b: int) -> int

Chains two runs of calculate_square and sums the results.

It uses ctx.run(...) to run each one on its own compute, and asyncio.gather(...) to run them in parallel:

result1, result2 = await asyncio.gather(
    ctx.run(calculate_square, a),
    ctx.run(calculate_square, b),
)

flip_coin(ctx) -> str

Simulates a coin flip:

• Heads: Returns success
• Tails: Raises an error to trigger retry

Retry policy in this example:

• max retries: 3
• wait duration: 1000ms
• backoff scaling: 1.5

Local Development

Prerequisites

• Python 3.10+

Run locally

Make sure you've installed the latest version of the Render CLI.

1. From this template's root, start the local task server:

   pip install -r requirements.txt
   render workflows dev -- python main.py

2. In a separate terminal, trigger task runs:

   render workflows tasks start calculate_square --local --input='{"a": 5}'
   render workflows tasks start sum_squares --local --input='{"a": 3, "b": 4}'
   render workflows tasks start flip_coin --local --input='{}'

Expected behavior:

• calculate_square with a=5 returns 25
• sum_squares with a=3,b=4 returns 25
• flip_coin may fail and retry before succeeding

Deploying to Render

Configure your Workflow service with:

Option	Value
Build command	pip install -r requirements.txt
Start command	python main.py

Key Concepts

Task registration

Any function decorated with @app.task is registered when your service starts via app.start().

Subtasks

Every task receives a TaskContext as its first parameter. Inside an async task, await ctx.run(other_task, ...) runs other_task on its own compute and returns its result. The context parameter isn't part of the task's inputs, so --input only supplies the parameters that follow it.

Retries

Use @app.task(retry=Retry(...)) when transient failures should be retried automatically.

Troubleshooting

"Task not found"

• Confirm the service is running
• Verify task names exactly match: calculate_square, sum_squares, flip_coin

Import or dependency issues

• Confirm dependency install completed from requirements.txt
• Confirm Python version is 3.10+

Resources

• Render Workflows documentation
• Workflows tutorial
• Local development guide