Protect Docker containers

The benefits of a Docker deployment are real, but so is the concern about the significant attack surface of the Docker host's operating system (OS) itself. Like any well-designed software deployment, OS hardening and the use of best practices for your deployment, such as the Center for Internet Security (CIS) Docker Benchmark, provide a solid foundation as a starting point. Once you have a secure foundation in place, adding Deep Security to your deployment gives you access to Trend Micro’s extensive experience protecting physical, virtual, and cloud workloads as well as to real-time threat information from the Trend Micro Smart Protection Network. Deep Security both protects your deployment as well as helps meet and maintain continuous compliance requirements. See Docker compatibility for information on supported Docker editions and releases.

Deep Security protects your Docker hosts and containers running on Linux distributions. Deep Security can do the following:

Deep Security Docker protection works at the OS level. This means that the Deep Security Agent must be installed on the Docker host's OS, not inside a container.

Communication between containers in the pod is not supported.

Beginning with Deep Security 10.1, Deep Security supports Docker in swarm mode while using an overlay network.

Deep Security protection for the Docker host

The following Deep Security modules can be used to protect the Docker host:

  • Intrusion Prevention (IPS)
  • Anti-Malware
  • Integrity Monitoring
  • Log Inspection
  • Application Control
  • Firewall
  • Web Reputation

Deep Security protection for Docker containers

The following Deep Security modules can be used to protect Docker containers:

  • Intrusion Prevention
  • Anti-Malware

Limitation on Intrusion Prevention recommendation scans

Although Deep Security Intrusion Prevention controls work at the host level, it also protects container traffic on the exposed container port numbers. Since Docker allows multiple applications to run on the same Docker host, a single Intrusion Prevention policy is applied to all Docker applications. This means that recommendation scans should not be relied upon for Docker deployments.