Infrastructure Services
Source index and context
crefft > Infrastructure > Services
Configuration guides for common infrastructure services. Each guide covers cross-platform setup with platform-specific sections where needed.
Service Guides
| Service | Guide | Status | Description |
|---|---|---|---|
| SSH | ssh.md | Complete | SSH server hardening and configuration |
| NTP | ntp.md | Draft | Time synchronization with chrony |
| DNS | dns.md | Draft | Resolver configuration, local DNS |
| Firewall | nftables.md | Draft | nftables firewall configuration |
| Logging | logging.md | Draft | Centralized logging and log rotation |
Draft means the guide has a Quick Reference table and section headings but
unwritten bodies marked TODO. Draft guides MUST NOT be treated as
normative until the TODOs are resolved.
TLS certificates (ACME/Let’s Encrypt) have no guide and none is drafted; see Roadmap.
Roadmap
| Service | Planned coverage |
|---|---|
| Certificates | TLS certificate issuance and renewal, ACME/Let’s Encrypt, internal CAs |
Common Patterns
Service Checklist
Every service deployment SHOULD address:
- Authentication: How is access controlled?
- Encryption: Is traffic encrypted in transit?
- Logging: Are security events logged?
- Monitoring: Are health checks configured?
- Backup: Is configuration backed up?
- Updates: How are security updates applied?
Configuration Management
- MUST use Ansible, Puppet, or similar for production config
- MUST version control all configuration
- SHOULD test changes in staging before production
- SHOULD use templates with environment-specific values
Documentation
Each service SHOULD have:
- Purpose and scope
- Upstream documentation link
- Configuration file locations
- Relevant log files
- Troubleshooting commands
See Also
- OS Fundamentals — Operating system concepts
- Linux Guide — Debian-specific configuration
- Ansible Guide — Automated service configuration