lmiranda
2d51df7a42
Phase 1b: Rename all ~94 commands across 12 plugins to /<noun> <action> sub-command pattern. Git-flow consolidated from 8→5 commands (commit variants absorbed into --push/--merge/--sync flags). Dispatch files, name: frontmatter, and cross-reference updates for all plugins. Phase 2: Design documents for 8 new plugins in docs/designs/. Phase 3: Scaffold 8 new plugins — saas-api-platform, saas-db-migrate, saas-react-platform, saas-test-pilot, data-seed, ops-release-manager, ops-deploy-pipeline, debug-mcp. Each with plugin.json, commands, agents, skills, README, and claude-md-integration. Marketplace grows from 12→20. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
128 lines
2.5 KiB
Markdown
128 lines
2.5 KiB
Markdown
# Docker Compose Patterns Skill
|
|
|
|
Best practices and patterns for Docker Compose service definitions targeting self-hosted infrastructure.
|
|
|
|
## Service Naming
|
|
|
|
- Use lowercase with hyphens: `my-service`
|
|
- Prefix with stack name for multi-project hosts: `myapp-db`, `myapp-redis`
|
|
- Container name should match service name: `container_name: myapp-db`
|
|
|
|
## Network Isolation
|
|
|
|
Every stack should define its own bridge network:
|
|
|
|
```yaml
|
|
networks:
|
|
app-network:
|
|
driver: bridge
|
|
```
|
|
|
|
Services join the stack network. Only the reverse proxy entry point should be exposed to the host.
|
|
|
|
## Volume Management
|
|
|
|
- Use **named volumes** for data persistence (databases, uploads)
|
|
- Use **bind mounts** for configuration files only
|
|
- Set explicit permissions with `:ro` for read-only mounts
|
|
- Label volumes with `labels` for identification
|
|
|
|
```yaml
|
|
volumes:
|
|
db-data:
|
|
labels:
|
|
com.project: myapp
|
|
com.service: database
|
|
```
|
|
|
|
## Healthchecks
|
|
|
|
Every service MUST have a healthcheck:
|
|
|
|
```yaml
|
|
healthcheck:
|
|
test: ["CMD", "curl", "-f", "http://localhost:8080/health"]
|
|
interval: 30s
|
|
timeout: 10s
|
|
retries: 3
|
|
start_period: 40s
|
|
```
|
|
|
|
Common healthcheck patterns:
|
|
- HTTP: `curl -f http://localhost:<port>/health`
|
|
- PostgreSQL: `pg_isready -U <user>`
|
|
- Redis: `redis-cli ping`
|
|
- MySQL: `mysqladmin ping -h localhost`
|
|
|
|
## Restart Policies
|
|
|
|
| Environment | Policy |
|
|
|-------------|--------|
|
|
| Development | `restart: "no"` |
|
|
| Production | `restart: unless-stopped` |
|
|
| Critical services | `restart: always` |
|
|
|
|
## Resource Limits
|
|
|
|
For Raspberry Pi (8GB RAM):
|
|
|
|
```yaml
|
|
deploy:
|
|
resources:
|
|
limits:
|
|
memory: 256M
|
|
cpus: '1.0'
|
|
reservations:
|
|
memory: 128M
|
|
```
|
|
|
|
## Dependency Ordering
|
|
|
|
Use healthcheck-aware dependencies:
|
|
|
|
```yaml
|
|
depends_on:
|
|
db:
|
|
condition: service_healthy
|
|
redis:
|
|
condition: service_started
|
|
```
|
|
|
|
## Environment Variables
|
|
|
|
Never inline secrets. Use `env_file`:
|
|
|
|
```yaml
|
|
env_file:
|
|
- .env
|
|
- .env.${DEPLOY_ENV:-development}
|
|
```
|
|
|
|
## Multi-Service Patterns
|
|
|
|
### Web App + Database + Cache
|
|
|
|
```yaml
|
|
services:
|
|
app:
|
|
image: myapp:1.0.0
|
|
env_file: [.env]
|
|
depends_on:
|
|
db: { condition: service_healthy }
|
|
redis: { condition: service_healthy }
|
|
networks: [app-network]
|
|
|
|
db:
|
|
image: postgres:16-alpine
|
|
volumes: [db-data:/var/lib/postgresql/data]
|
|
healthcheck:
|
|
test: ["CMD", "pg_isready", "-U", "postgres"]
|
|
networks: [app-network]
|
|
|
|
redis:
|
|
image: redis:7-alpine
|
|
healthcheck:
|
|
test: ["CMD", "redis-cli", "ping"]
|
|
networks: [app-network]
|
|
```
|