Skip to content

First service: config → job store → Job → Worker → COMPLETE

This tutorial is a single narrative path with expected outcomes and common pitfalls. For install commands and the full example list, see Getting Started.

Goal

From a clean checkout, run the scheduler example so a job reaches Completed in the in-memory job store.

Prerequisites

  • Python 3.10–3.14 and repo dependencies (make build-prepare or uv pip install -r requirements-dev.txt then requirements.txt)
  • Work from the repository root (config and example paths are relative)

Step 1 — Load config and wire the job store

config.load_config() applies log/tracing settings. It does not configure a durable scheduler.jobStore. Always call setup_from_config after load when YAML declares a store (even memory).

from pypepper.common.config import config
from pypepper.scheduler.store import setup_from_config

config.load_config("./conf/app.config.yaml")
setup_from_config(config.get_yml_config())

Forgot setup_from_config?

A non-memory scheduler.jobStore.backend in YAML makes Job.save / Job.get_saved raise ValueError until you call setup_from_config (or configure_job_store / set_job_store). Declaring backend: memory does not arm that guard.

Step 2 — Run the scheduler example

python example/scheduler/app.py

What it does:

  1. Builds a one-task workflow with CallableExecutor
  2. Calls Job.scheduled() in a sync context (required — no running event loop)
  3. Runs Worker.run_once() until the job finishes

Expected outcome

Logs should include a line that the job completed, for example:

job completed: id=... status=Completed

The process exits 0. If an assertion fails inside the example, you will see a traceback instead.

Called scheduled() from async code?

Job.scheduled() / Processor.run raise RuntimeError when an event loop is already running. Keep scheduling in sync code (as the example does), then asyncio.run only the worker.

Step 3 (optional) — SSE echo stream

In another terminal:

export PYPEPPER_SSE_API_KEY=your-local-key
python example/sse/app.py

Connect:

curl -N -H "X-API-Key: $PYPEPPER_SSE_API_KEY" http://localhost:55550/sse/echo

You should see SSE event frames. Keep sse.authentication.enabled: true outside local experiments. Auth-off requires PYPEPPER_SSE_ALLOW_AUTH_OFF=1 (see SSE security notes).

Next