mirror of
https://github.com/x1xhlol/system-prompts-and-models-of-ai-tools.git
synced 2026-06-18 07:19:35 +00:00
182 lines
4.0 KiB
Markdown
182 lines
4.0 KiB
Markdown
# Deployment Guide
|
|
|
|
## 🐳 Docker (recommended for most teams)
|
|
|
|
### Single container (app only)
|
|
|
|
```bash
|
|
docker build -t ai-company-saudi:2.0.0 .
|
|
docker run -d \
|
|
--name ai-company \
|
|
--env-file .env \
|
|
-p 8000:8000 \
|
|
--restart unless-stopped \
|
|
ai-company-saudi:2.0.0
|
|
```
|
|
|
|
### Full stack with docker-compose
|
|
|
|
```bash
|
|
make docker-up
|
|
# app + PostgreSQL 16 + Redis 7 + MongoDB 7
|
|
```
|
|
|
|
Stop:
|
|
```bash
|
|
make docker-down
|
|
```
|
|
|
|
Logs:
|
|
```bash
|
|
make docker-logs
|
|
```
|
|
|
|
---
|
|
|
|
## ☁️ Production VPS (bare metal / DigitalOcean / Hetzner / AWS EC2)
|
|
|
|
### 1. Prep the server
|
|
|
|
```bash
|
|
# Debian/Ubuntu
|
|
sudo apt update && sudo apt install -y python3.12 python3.12-venv git nginx postgresql redis
|
|
```
|
|
|
|
### 2. Clone & configure
|
|
|
|
```bash
|
|
git clone https://github.com/YOUR-ORG/ai-company-saudi.git /opt/ai-company
|
|
cd /opt/ai-company
|
|
|
|
python3.12 -m venv venv
|
|
source venv/bin/activate
|
|
pip install -e .
|
|
|
|
cp .env.example .env
|
|
# edit .env with your secrets
|
|
```
|
|
|
|
### 3. Systemd service
|
|
|
|
`/etc/systemd/system/ai-company.service`:
|
|
|
|
```ini
|
|
[Unit]
|
|
Description=AI Company Saudi API
|
|
After=network.target postgresql.service redis.service
|
|
|
|
[Service]
|
|
Type=exec
|
|
User=appuser
|
|
WorkingDirectory=/opt/ai-company
|
|
Environment="PATH=/opt/ai-company/venv/bin"
|
|
EnvironmentFile=/opt/ai-company/.env
|
|
ExecStart=/opt/ai-company/venv/bin/uvicorn api.main:app --host 127.0.0.1 --port 8000 --workers 4
|
|
Restart=on-failure
|
|
RestartSec=5s
|
|
|
|
[Install]
|
|
WantedBy=multi-user.target
|
|
```
|
|
|
|
Enable:
|
|
```bash
|
|
sudo systemctl daemon-reload
|
|
sudo systemctl enable --now ai-company
|
|
sudo systemctl status ai-company
|
|
```
|
|
|
|
### 4. nginx reverse proxy
|
|
|
|
`/etc/nginx/sites-available/ai-company`:
|
|
|
|
```nginx
|
|
server {
|
|
listen 80;
|
|
server_name api.ai-company.sa;
|
|
|
|
location / {
|
|
proxy_pass http://127.0.0.1:8000;
|
|
proxy_set_header Host $host;
|
|
proxy_set_header X-Real-IP $remote_addr;
|
|
proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
|
|
proxy_set_header X-Forwarded-Proto $scheme;
|
|
proxy_read_timeout 120s;
|
|
}
|
|
}
|
|
```
|
|
|
|
Enable + TLS:
|
|
```bash
|
|
sudo ln -s /etc/nginx/sites-available/ai-company /etc/nginx/sites-enabled/
|
|
sudo nginx -t
|
|
sudo systemctl reload nginx
|
|
|
|
# Let's Encrypt (certbot)
|
|
sudo certbot --nginx -d api.ai-company.sa
|
|
```
|
|
|
|
### 5. Database init
|
|
|
|
```bash
|
|
sudo -u postgres psql -c "CREATE USER ai_user WITH PASSWORD 'strong-password';"
|
|
sudo -u postgres psql -c "CREATE DATABASE ai_company OWNER ai_user;"
|
|
make db-init
|
|
```
|
|
|
|
---
|
|
|
|
## ☁️ Kubernetes (advanced)
|
|
|
|
Basic Deployment + Service manifests are available in `docs/k8s/` (coming). Use the published Docker image from GHCR:
|
|
|
|
```
|
|
ghcr.io/YOUR-ORG/ai-company-saudi:v2.0.0
|
|
```
|
|
|
|
Essentials:
|
|
- Use a **Secret** for `.env` contents
|
|
- Use a **ConfigMap** for non-sensitive settings
|
|
- Set up liveness (`/live`), readiness (`/ready`), and startup probes
|
|
- Request 500m CPU / 512Mi memory; limit 2 CPU / 2Gi memory
|
|
|
|
---
|
|
|
|
## ☁️ Managed Python platforms
|
|
|
|
### Railway / Render / Fly.io
|
|
|
|
All three support auto-deploy from GitHub. Steps:
|
|
1. Create a new app, connect this repo.
|
|
2. Add all env vars from `.env.example` as secrets.
|
|
3. Set start command: `uvicorn api.main:app --host 0.0.0.0 --port $PORT --workers 2`.
|
|
4. Attach a managed PostgreSQL add-on, copy its URL to `DATABASE_URL`.
|
|
|
|
---
|
|
|
|
## 📊 Health monitoring
|
|
|
|
Every deployment should configure:
|
|
- **Uptime checks** against `/health` (30s interval)
|
|
- **Alerts** on:
|
|
- `/health` returning non-200 for > 2 minutes
|
|
- Error rate > 1% over 5 minutes
|
|
- p95 latency > 10s for > 5 minutes
|
|
- **Log aggregation** — ship structlog JSON to Grafana Loki, Datadog, or Elastic
|
|
|
|
---
|
|
|
|
## 🔐 Production hardening checklist
|
|
|
|
- [ ] `.env` has real keys, `APP_ENV=production`, `APP_DEBUG=false`
|
|
- [ ] `APP_SECRET_KEY` is a 64-char random string
|
|
- [ ] `CORS_ORIGINS` restricted to your actual domains
|
|
- [ ] Rate limiting in nginx / Cloudflare
|
|
- [ ] TLS certificate valid
|
|
- [ ] Database credentials strong + rotated quarterly
|
|
- [ ] Backup schedule for Postgres (daily, 30-day retention)
|
|
- [ ] Dependabot alerts enabled on GitHub
|
|
- [ ] Branch protection on `main`
|
|
- [ ] Required reviews on PRs
|
|
- [ ] No `.env` committed (run `gitleaks detect --source .`)
|