Skip to main content

Server Setup

Overview​

Defines the base configuration and setup for all server instances in the infrastructure. This includes both production and development servers, with specific considerations for different roles and requirements.

Status​

  • Base server setup with essential packages
  • Container runtime management (Docker/Containerd)
  • Logrotate and unattended upgrades
  • Network monitoring tools
  • Monitoring integration (Prometheus/Node Exporter)
  • Backup configuration
  • Resource limits and tuning

Components​

Base Setup​

  • Essential packages installation
  • System maintenance tools
  • Network monitoring utilities
  • Log management
  • Automatic updates

Container Runtime​

  • Docker for standalone services (dockerio role)
  • Containerd for Kubernetes nodes (containerd role)
  • No default container runtime installation

Server Types​

  1. Production Servers

    • Mercury TA servers
    • Mercury Bot servers
    • Kubernetes nodes
    • Dokku servers
    • GPU/CPU compute servers

  2. Development Servers

    • VPN servers
    • Test environments
    • Development proxies

Technical Implementation​

Playbooks​

  • server_setup.yml: Base server configuration
  • production_setup.yml: Production-specific setup with container runtime handling

Roles​

  • server: Base server configuration
  • docker: Docker installation and configuration (when dockerio role is present)
  • containerd: Containerd setup for Kubernetes nodes
  • Various service-specific roles (dokku, monitoring, etc.)

Configuration​

# Server essentials
- net-tools
- nmap
- tcpdump
- iftop
- mtr
- rsync
- logrotate
- cron
- unattended-upgrades

# Container runtime selection via roles
roles:
  - dockerio     # For Docker-based servers
  - containerd   # For Kubernetes nodes

Development Guidelines​

Adding New Server Types​

  1. Define required roles in inventory
  2. Add specific configuration if needed
  3. Update relevant playbooks
  4. Document in FDD

Container Runtime Selection​

  • Use dockerio role for Docker-based services
  • Use containerd role for Kubernetes nodes
  • Never install both on the same host
  • Document runtime choice in inventory

Monitoring and Maintenance​

  • All servers should have basic monitoring
  • Regular backup configuration
  • Resource usage monitoring
  • Log rotation and management

Next Steps​

  1. Create monitoring role with Prometheus/Node Exporter
  2. Implement backup strategy
  3. Add resource monitoring and alerts
  4. Create server hardening role
  5. Implement automatic certificate management