Why containerise internal tooling?
MSPs run a long tail of internal tools — monitoring dashboards, ticketing, n8n, NextCloud, password vaults. Each is a dependency and upgrade headache on bare metal. Containerising them means a tool is a docker-compose file plus a volume: reproducible, version-pinned, and rebuildable on a new host in minutes.
Compose for multi-service stacks
Docker Compose defines the whole stack — app, database, reverse proxy — in one declarative file. Use named volumes for persistent data, a dedicated bridge network for inter-service traffic, and an external proxy (Caddy or Traefik) for automatic TLS termination.
- Pin image tags to a digest, not "latest"
- Use named volumes for data that must survive rebuilds
- Put a reverse proxy in front for TLS and routing
- Keep compose files in Git alongside the runbook
Operational concerns
Back up the volumes, not just the containers — the container is disposable, the data is not. Schedule image rebuilds on a cadence so security patches land. Monitor container health with a healthcheck and let the orchestrator restart unhealthy containers.
Lessons learned
- Pin image versions — "latest" breaks at the worst time
- Back up volumes, not containers
- Use healthchecks so unhealthy containers auto-restart
- Commit compose files to Git so the stack is reproducible
Written by
Jamshed Alam
Global MSP & IT Infrastructure Consultant with 20 years across Exchange, Microsoft 365, Azure, security and automation. Available for project and contract work worldwide.