Process Orchestration

Process Orchestration is pyserve's flagship feature for running multiple Python web applications with full process isolation. Each application runs in its own subprocess with independent lifecycle, health monitoring, and automatic restart on failure.

Overview

Unlike ASGI Mounting which runs apps in-process, Process Orchestration provides:

Process Isolation — Each app runs in a separate Python process
Health Monitoring — Automatic health checks with configurable intervals
Auto-restart — Failed processes restart with exponential backoff
Multi-worker Support — Configure multiple uvicorn workers per app
Dynamic Port Allocation — Automatic port assignment (9000-9999)
WSGI Support — Flask/Django apps via automatic wrapping
Request Tracing — X-Request-ID propagation through proxied requests

Architecture


                    PyServe Gateway (:8000)
                           │
          ┌────────────────┼────────────────┐
          ▼                ▼                ▼
      FastAPI          Flask           Starlette
       :9001           :9002             :9003
      /api/*          /admin/*           /ws/*

PyServe acts as a gateway, routing requests to the appropriate subprocess based on URL path.

Basic Configuration

server:
  host: 0.0.0.0
  port: 8000

extensions:
  - type: process_orchestration
    config:
      apps:
        - name: api
          path: /api
          app_path: myapp.api:app
          
        - name: admin
          path: /admin
          app_path: myapp.admin:app

App Configuration Options

name: Unique identifier for the application (required)
path: URL path prefix for routing requests (required)
app_path: Python import path. Format: module:attribute (required)
app_type: Application type: asgi or wsgi. Default: asgi
workers: Number of uvicorn workers. Default: 1
port: Fixed port number. Default: auto-allocated from port_range
factory: If true, app_path points to a factory function. Default: false
env: Environment variables to pass to the subprocess
module_path: Path to add to sys.path for module resolution

Health Check Options

health_check_enabled: Enable health monitoring. Default: true
health_check_path: Endpoint to check for health. Default: /health
health_check_interval: Interval between health checks in seconds. Default: 10.0
health_check_timeout: Timeout for health check requests. Default: 5.0
health_check_retries: Failed checks before restart. Default: 3

Restart Options

max_restart_count: Maximum restart attempts before giving up. Default: 5
restart_delay: Initial delay between restarts in seconds. Default: 1.0
shutdown_timeout: Timeout for graceful shutdown. Default: 30.0

Global Configuration

extensions:
  - type: process_orchestration
    config:
      port_range: [9000, 9999]
      health_check_enabled: true
      proxy_timeout: 60.0
      logging:
        httpx_level: warning
        proxy_logs: true
        health_check_logs: false
      apps:
        # ...

port_range: Range for dynamic port allocation. Default: [9000, 9999]
proxy_timeout: Timeout for proxied requests in seconds. Default: 60.0
logging.httpx_level: Log level for HTTP client (debug/info/warning/error). Default: warning
logging.proxy_logs: Log proxied requests with latency. Default: true
logging.health_check_logs: Log health check results. Default: false

FastAPI Example

# myapp/api.py
from fastapi import FastAPI

app = FastAPI()

@app.get("/health")
async def health():
    return {"status": "ok"}

@app.get("/users")
async def get_users():
    return [{"id": 1, "name": "Alice"}]

extensions:
  - type: process_orchestration
    config:
      apps:
        - name: api
          path: /api
          app_path: myapp.api:app
          workers: 4
          health_check_path: /health

Requests to /api/users are proxied to the FastAPI process as /users.

Flask Example (WSGI)

# myapp/admin.py
from flask import Flask

app = Flask(__name__)

@app.route("/health")
def health():
    return {"status": "ok"}

@app.route("/dashboard")
def dashboard():
    return {"page": "dashboard"}

extensions:
  - type: process_orchestration
    config:
      apps:
        - name: admin
          path: /admin
          app_path: myapp.admin:app
          app_type: wsgi
          workers: 2

Note: WSGI support requires a2wsgi package: pip install a2wsgi

Factory Pattern

# myapp/api.py
from fastapi import FastAPI

def create_app(debug: bool = False) -> FastAPI:
    app = FastAPI(debug=debug)
    
    @app.get("/health")
    async def health():
        return {"status": "ok", "debug": debug}
    
    return app

apps:
  - name: api
    path: /api
    app_path: myapp.api:create_app
    factory: true

Environment Variables

Pass environment variables to subprocesses:

apps:
  - name: api
    path: /api
    app_path: myapp.api:app
    env:
      DATABASE_URL: "postgresql://localhost/mydb"
      REDIS_URL: "redis://localhost:6379"
      DEBUG: "false"

Multiple Applications

extensions:
  - type: process_orchestration
    config:
      port_range: [9000, 9999]
      apps:
        # FastAPI REST API
        - name: api
          path: /api
          app_path: apps.api:app
          workers: 4
        
        # Flask Admin Panel
        - name: admin
          path: /admin
          app_path: apps.admin:app
          app_type: wsgi
          workers: 2
        
        # Starlette WebSocket Handler
        - name: websocket
          path: /ws
          app_path: apps.websocket:app
          workers: 1

Request Tracing

PyServe automatically generates and propagates X-Request-ID headers:

If a request has X-Request-ID, it's preserved
Otherwise, a UUID is generated
The ID is passed to subprocesses and included in response headers
All logs include the request ID for tracing

Process Orchestration vs ASGI Mount

Feature	Process Orchestration	ASGI Mount
Isolation	Full process isolation	Shared process
Memory	Separate per app	Shared
Crash Impact	Only that app restarts	All apps affected
Health Checks	Yes, with auto-restart	No
Multi-worker	Yes, per app	No
Latency	HTTP proxy overhead	In-process (faster)
Use Case	Production, isolation needed	Development, simple setups

When to use Process Orchestration:

Running multiple apps that shouldn't affect each other
Need automatic restart on failure
Different resource requirements per app
Production deployments

When to use ASGI Mount:

Development and testing
Simple setups with trusted apps
Minimal latency requirements

See Also:

ASGI Mounting Guide — In-process app mounting
Extensions Reference — All extension types
Configuration Guide — Full configuration reference