Our infrastructure team is responsible for building and operating separate Docker environments for development, testing, and production, but we are not responsible for application development or the broader DevOps workflow. We understand the basics from experimenting with Docker, but we need deeper training focused on the server-management side: securing container networks, configuring reverse proxies, managing hosts, integrating image repositories, controlling how developers deploy images, and establishing sound operational practices. Most courses we've found focus heavily on building applications and CI/CD pipelines, while giving little attention to securely managing the underlying Docker infrastructure. What training or learning resources would you recommend for systems administrators in this situation?
4 Answers
Hands-on practice is probably the fastest way to fill the gaps, but make the lab resemble production. Run separate environments, use private images, restrict container-to-container traffic, test what happens when a container is compromised, and document the process for developers to publish and deploy images. That will teach the management concerns that introductory Docker courses often skip.
KodeKloud has Docker-focused training that may be a useful starting point. For a more practical approach, build a small lab and work through host hardening, image and registry access, network isolation, reverse proxies, storage, logging, backups, and upgrades rather than only following application-development tutorials.
There is some unavoidable overlap with DevOps because Docker affects deployment, networking, and operational workflows, but that doesn’t mean infrastructure administrators need to become application developers. Focus your study on Docker Engine and host security, least-privilege access, exposed ports, container capabilities, registry authentication, patching, resource limits, and how the reverse proxy and firewall are configured.
Portainer is worth evaluating if you want a management interface with fairly deep visibility and integrations, including Git-based workflows. Coolify is another option for making deployments more hands-off, although you should still understand the Docker hosts and security model underneath instead of treating the platform as a replacement for that knowledge.
Experimenting alone doesn’t guarantee good security practices, so I’d pair the lab work with Docker’s official security guidance and material covering Linux isolation, namespaces, capabilities, registries, and reverse-proxy design.