Real User Survey

Simple Django application that captures qualitative survey responses for the getsva research initiative.

Local setup

cd /Users/anuragsingh/Documents/GitHub/survey
python3 -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt
python manage.py migrate
python manage.py runserver

Visit http://127.0.0.1:8000 to fill out the form.

Running tests

source .venv/bin/activate
python manage.py test

Production deployment on Cloudflare Workers (survey.getsva.com)

Cloudflare’s Python Workers runtime can host this Django app as long as the database lives in a remotely reachable service (Neon, Supabase, RDS, etc.). The checklist below covers everything you need to do locally before the deploy, followed by the exact steps to run inside Cloudflare.

Local preparation (one-time)

1. Create env files

   cd /Users/anuragsingh/Documents/GitHub/survey
   cp env.example env.local
   cp env.example env.production

   Fill in:

   • ENVIRONMENT=local (or production inside env.production)
   • SECRET_KEY= new random string
   • DEBUG=false in prod
   • ALLOWED_HOSTS=survey.getsva.com,*.workers.dev
   • CSRF_TRUSTED_ORIGINS=https://survey.getsva.com,https://<your-worker>.workers.dev
   • DATABASE_URL= connection string for your managed cloud DB (e.g., postgres://user:pass@host:5432/dbname)
   • Any email/S3 keys you need later.

2. Update Django settings

   • In survey_site/settings.py, load env vars (e.g., using os.environ or django-environ) and configure:
     • SECRET_KEY = os.environ["SECRET_KEY"]
     • DEBUG = os.environ.get("DEBUG", "false").lower() == "true"
     • ALLOWED_HOSTS = os.environ["ALLOWED_HOSTS"].split(",")
     • CSRF_TRUSTED_ORIGINS = os.environ["CSRF_TRUSTED_ORIGINS"].split(",")
     • DATABASES["default"] from DATABASE_URL when present (fall back to SQLite for local dev).
   • Set STATIC_ROOT = BASE_DIR / "staticfiles" so collectstatic has a deterministic output.

3. Expose Django as ASGI

   • Ensure survey_site/__init__.py sets DJANGO_SETTINGS_MODULE.
   • Add worker.py at the repo root:

     import os
     import django
     from asgiref.wsgi import WsgiToAsgi
     from django.core.wsgi import get_wsgi_application
     
     os.environ.setdefault("DJANGO_SETTINGS_MODULE", "survey_site.settings")
     django.setup()
     asgi_app = WsgiToAsgi(get_wsgi_application())

4. Create the Worker project

   npm install -g wrangler
   wrangler init survey-worker --python --no-config
   mv worker.py functions

   Structure after this step (key files only):

   .
   ├─ functions/
   │  └─ _worker.py
   ├─ survey_site/
   ├─ surveys/
   ├─ worker.py
   ├─ wrangler.toml
   

5. Wire the Worker entry point

   • Replace functions/_worker.py with:

     from worker import asgi_app
     
     async def on_fetch(request, env, ctx):
         return await asgi_app(request, env, ctx)

   • Set up wrangler.toml:

     name = "survey-getSVA"
     main = "functions/_worker.py"
     compatibility_date = "2025-11-17"
     compatibility_flags = ["python_workers"]
     
     [vars]
     DJANGO_SETTINGS_MODULE = "survey_site.settings"
     SECRET_KEY = "copy-from-.env.production"
     ALLOWED_HOSTS = "survey.getsva.com,*.workers.dev"
     CSRF_TRUSTED_ORIGINS = "https://survey.getsva.com,https://survey-getSVA.workers.dev"
     DATABASE_URL = "postgres://..."

6. Install dependencies & freeze

   python -m pip install -r requirements.txt
   pip freeze > requirements.txt  # ensures Worker build sees exact versions

7. Static assets

   python manage.py collectstatic --settings=survey_site.settings --noinput

   Upload the resulting staticfiles/ directory to Cloudflare R2 (recommended) or keep it alongside the Worker and serve via WhiteNoise if the footprint is small.

8. Local validation

   python manage.py migrate --settings=survey_site.settings
   python manage.py test

   Run against your managed DB (temporarily allow your IP if needed) to confirm migrations succeed before the Worker touches production data.

Commit all necessary files (worker.py, functions/_worker.py, wrangler.toml, updated settings) and push so Wrangler can deploy from the clean state.

Actions inside Cloudflare

1. Authenticate Wrangler

   cd /Users/anuragsingh/Documents/GitHub/survey
   wrangler login

2. Deploy the Worker

   wrangler deploy
   wrangler tail  # stream logs for sanity check

   Ensure your external database allows connections from Cloudflare’s egress IPs (most managed Postgres/Supabase instances permit this by default).

3. Run migrations in the Worker environment

   wrangler invoke survey-getSVA --command "python manage.py migrate --settings=survey_site.settings"

4. Bind secrets (optional) For values you do not want inside wrangler.toml, use:

   wrangler secret put SECRET_KEY
   wrangler secret put DATABASE_URL

   and remove them from the [vars] block.

5. Attach the custom domain

   • Cloudflare Dashboard → Workers & Pages → survey-getSVA → Triggers → Add Custom Domain → survey.getsva.com.
   • Accept the suggested CNAME DNS record; keep the proxy enabled (orange cloud).
   • Enable “Always use HTTPS” under SSL/TLS → Edge Certificates.

6. Cache controls

   • Create a Cache Rule: if path starts_with "/admin" or Request Method is POST → Cache level: Bypass.
   • Optional rule to cache static assets aggressively if served via Worker/R2.

7. Smoke test

   • Visit the workers.dev preview URL to ensure forms work and data lands in the DB.
   • After DNS propagates, verify https://survey.getsva.com for both survey form and /admin.
   • Monitor logs via wrangler tail during initial traffic.

Once these steps are complete, the Django app runs entirely inside Cloudflare’s Python Workers runtime while persisting data to your managed cloud database.