Deployment

Standard Deployment

pymode deploy

This bundles your Python code, stdlib, and dependencies into a Cloudflare Worker with a PythonDO Durable Object.

What Gets Deployed

Component	Size	Description
python.wasm	5.7MB (1.8MB gz)	CPython 3.13 runtime
stdlib bundle	~150KB	Essential stdlib modules
Your code	varies	Python source files
Dependencies	varies	Installed packages
Worker JS	~20KB	Worker entry + WASI + Asyncify runtime

Total bundle size is typically 2-3MB gzipped, well within Cloudflare’s 10MB limit.

Wizer Snapshots (Experimental)

Wizer pre-initializes the Python runtime at deploy time, reducing cold start from ~28ms to ~5ms. The build script exists (scripts/build-wizer.py) but this feature requires additional testing.

pymode deploy --wizer

How It Works

1. At deploy time, Wizer runs python.wasm up to the point where CPython is fully initialized
2. It takes a memory snapshot at that point
3. At runtime, the snapshot is loaded directly — skipping all initialization

This is particularly effective because CPython’s startup (loading frozen modules, initializing builtins, importing site) is the dominant cold-start cost.

Environment Variables and Secrets

Development

Create a .dev.vars file in your project root:

API_KEY=sk-test-123
DATABASE_URL=postgres://localhost:5432/mydb

Production

Use Wrangler to set secrets:

# Set a secret
wrangler secret put API_KEY

# List secrets
wrangler secret list

# Delete a secret
wrangler secret delete API_KEY

Access in Python:

def on_fetch(request, env):
    api_key = env.API_KEY

Custom Domains

Configure in wrangler.toml:

routes = [
  { pattern = "api.example.com/*", zone_name = "example.com" }
]

Or use Cloudflare’s dashboard to add a custom domain to your Worker.

Monitoring

Logs

View real-time logs:

wrangler tail

Metrics

Cloudflare Workers analytics are available in the dashboard:

• Request count
• CPU time per request
• Error rate
• Cold start frequency

Error Tracking

Python tracebacks are captured in Worker logs. Use wrangler tail or the Cloudflare dashboard to view them.

def on_fetch(request, env):
    try:
        return handle(request, env)
    except Exception as e:
        # Logged automatically to Worker logs
        print(f"Error: {e}", file=sys.stderr)
        return Response("Internal Server Error", status=500)

Limits

Resource	Limit
Worker size	10MB (compressed)
CPU time per request	30s
Memory per DO	128MB
Subrequests per request	50
KV value size	25MB
R2 object size	5GB
D1 database size	10GB

CI/CD

GitHub Actions

name: Deploy

on:
  push:
    branches: [main]

jobs:
  deploy:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4

      - uses: actions/setup-node@v4
        with:
          node-version: "20"

      - name: Install dependencies
        run: npm install

      - name: Deploy
        run: npx wrangler deploy
        env:
          CLOUDFLARE_API_TOKEN: ${{ secrets.CF_API_TOKEN }}

Required Secrets

Secret	Description
CF_API_TOKEN	Cloudflare API token with Workers write permission