SF Platform Documentation | SF Platform Documentation

On this page

% Environment & Deployment Runbook

Last updated: 2025-11-08

Environment Model

Environment	Description	Base URL	Notes
Development	Local developer machines	`http://localhost:3000`	Uses `.env.local`; no auth restrictions.
Staging	Pre-production validation	`https://staging.api.example.com`	Mirrors production data shape; limited to staging accounts.
Production	Live tenant	`https://api.example.com`	Requires elevated permissions and change control.

Configure additional environments in src/config/environments.ts. Each entry needs name, apiUrl, wsUrl, and optional displayName.

Environment Variables

Create .env.local for local dev. Example:

VITE_API_URL=http://localhost:3000
VITE_WS_URL=ws://localhost:3000
VITE_APP_TITLE=SF Middleware Dashboard
VITE_APP_VERSION=1.0.0
VITE_SENTRY_DSN=

Do not commit .env.local.
Use .env.staging / .env.production for build pipelines.
Document new variables in this file and README.

Deployment Paths

Static Hosting

npm install
npm run build
Upload dist/ to hosting (S3, Netlify, Azure Static Web Apps, etc.).
Ensure rewrites route all paths to index.html (for client-side routing).

Docker Image

docker build -t sf-dashboard .
docker run -p 3001:3001 sf-dashboard

Use multi-stage Dockerfile (build + nginx) to keep image small.
Inject environment variables at runtime via env.js or server config.

CI/CD Checklist

Lint & type check (npm run lint, npm run typecheck if configured).
Build test (npm run build).
Bundle analysis (optional).
Deploy to staging; run smoke suite.
Promote to production after approval.

Operational Runbook

Smoke Tests (post-deploy):

Visit /overview; confirm metrics load without console errors.
Login/logout flow works.
Environment selector switches between at least two environments.
Execute sample export from /audit.

Rollback Strategy:

Static hosting: revert to previous build artifact; ensure caches invalidated.
Docker: redeploy previous tagged image.
Document cause and remediation steps in incident log.

Monitoring Signals:

Frontend availability (uptime monitor).
API latency/error rate (APM dashboards).
Export job success metrics.
WebSocket connection failure rates (if enabled).

Change Management

Schedule production releases during maintenance windows.
Announce user-visible changes in docs/CHANGELOG.md.
Update documentation alongside code changes; require PR checklist confirmation.