Denozordec
ed4d6f8a46
quality / commitlint (push) Skipped
CD / update-wiki (push) Failing after 8s
quality / changes (push) Successful in 7s
quality / docker-check (push) Skipped
quality / web (push) Successful in 1m8s
quality / api (push) Successful in 41s
CD / quality (push) Successful in 2m0s
CD / publish (push) Failing after 11m24s
- Removed Dockerfile and Dockerfile.test as part of the transition to a pre-built image. - Updated .dockerignore and .gitignore to reflect new build context and ignored files. - Modified docker-compose.yml to use the new image for the application. - Enhanced documentation in AGENTS.md and README.md to include CI/CD details and release process. - Added new dependencies for commit linting and semantic release in package.json and pnpm-lock.yaml. This commit streamlines the Docker setup and improves the CI/CD workflow with Gitea Actions.
82 lines
3.6 KiB
Markdown
82 lines
3.6 KiB
Markdown
# Cloudflare Domain Manager
|
||
|
||
Self-hosted service for managing domains, DNS records, and SSL certificates via Cloudflare API.
|
||
|
||
## Features
|
||
|
||
- Domain and DNS record management (hybrid sync with Cloudflare)
|
||
- Domain groups (local / vpn / external) and service tags
|
||
- Service groups with common FQDN and DNS-based load balancing (Round Robin / Failover / Weighted)
|
||
- Health checks (TCP / HTTP) with automatic DNS reconciliation on status change
|
||
- SSL certificate expiry monitoring
|
||
- React UI with TanStack Router & Query
|
||
- Single Docker container deployment
|
||
|
||
## Quick start
|
||
|
||
```bash
|
||
cp .env.example .env
|
||
# Edit CLOUDFLARE_API_TOKEN, JWT_SECRET
|
||
|
||
docker compose up -d
|
||
```
|
||
|
||
Open http://localhost:8080 — default login `admin` / `admin` (dev only).
|
||
|
||
## Stack
|
||
|
||
- **Backend:** Node.js 22, Fastify 5, Drizzle ORM, SQLite (`apps/api`)
|
||
- **Shared:** Zod schemas (`packages/shared`), DB layer (`packages/db`)
|
||
- **Frontend:** pnpm monorepo (`apps/web` + `packages/ui`), React, Vite, TanStack Router/Query/Table, shadcn/ui (base-nova), Recharts
|
||
- **CI:** Gitea Actions ([docs/releasing.md](docs/releasing.md)) — quality gates + semantic-release + bake (`cfdm` / `cloudflare-domain-manager`)
|
||
|
||
## Development
|
||
|
||
```bash
|
||
pnpm install
|
||
pnpm --filter @cfdm/shared build
|
||
pnpm --filter @cfdm/db build
|
||
|
||
# API + SQLite (:8080)
|
||
pnpm --filter @cfdm/api dev
|
||
|
||
# Frontend (:5173, proxies /api → :8080)
|
||
pnpm --filter web dev
|
||
```
|
||
|
||
```bash
|
||
pnpm turbo build
|
||
pnpm turbo test
|
||
```
|
||
|
||
shadcn CLI: `cd apps/web && pnpm dlx shadcn@latest add <component>`
|
||
|
||
## Documentation
|
||
|
||
See [docs/Home.md](docs/Home.md), [docs/releasing.md](docs/releasing.md) and [CONTRIBUTING.md](CONTRIBUTING.md).
|
||
|
||
## Load balancing & health checks
|
||
|
||
Группа сервисов может иметь общий домен (`service_groups.domain`). На общем домене и на
|
||
привязках сервиса с несколькими A-записями включается балансировка и health-check:
|
||
|
||
- **Режимы LB:** `round_robin` (по одной A на каждый up-IP), `failover` (A только для
|
||
up-IP с минимальным приоритетом; бэкапы подключаются при падении primary),
|
||
`weighted` (веса учитываются в БД; в Cloudflare free отображается как Round Robin, т.к.
|
||
CF API не допускает дубликаты `(name, type, content)` A-записей — для истинного weighted
|
||
нужен CF Load Balancer).
|
||
- **Health-check:** TCP connect или HTTP (настраиваемый порт, путь, ожидаемый статус,
|
||
интервал, таймаут). Результаты хранятся в `ip_health_status` (`up` / `degraded` / `down`
|
||
/ `unknown`) и опрашиваются UI с polling 10с.
|
||
- **Reconcile:** при смене статуса IP cron перезаписывает A-записи в Cloudflare, оставляя
|
||
только активные по политике LB. Реакция = TTL A-записи (`ttl=1` proxied — минимальный).
|
||
|
||
Env для health-check:
|
||
|
||
| Variable | Default | Description |
|
||
|----------|---------|-------------|
|
||
| `HEALTH_CHECK_CRON` | `*/30 * * * * *` | Cron выражение для запуска проверок (каждые 30с) |
|
||
| `HEALTH_DEGRADED_FAILURES` | `1` | Порог последовательных ошибок → статус `degraded` |
|
||
| `HEALTH_DOWN_FAILURES` | `2` | Порог последовательных ошибок → статус `down` (IP убирается из DNS) |
|
||
| `HEALTH_LATENCY_WARN_MS` | `1000` | Латентность выше порога → `degraded` даже при успешном connect |
|