Runtime Security
When to Use This Skill
This section covers runtime security for GKE clusters:
- Pod Security Standards: Namespace-level security policies (baseline, restricted)
- Admission Controllers: Pre-deployment validation and policy enforcement
- Runtime Monitoring: Behavioral detection with Falco or GKE Cloud Logging
Prerequisites
- GCP project with billing enabled
- Terraform 1.0+
- kubectl configured for cluster access
Implementation
- Cluster Configuration - Private GKE, Workload Identity
- Network Security - VPC networking, Network Policies
- IAM Configuration - Least-privilege IAM
Key Principles
Defense in Depth
Multiple layers of runtime security controls:
- Pod Security Standards enforce secure defaults
- Admission controllers block invalid configurations
- Runtime monitoring detects anomalous behavior
- Audit logging captures all activity
Secure by Default
Production workloads must meet strict security requirements:
- Run as non-root user
- Read-only root filesystem
- Drop all Linux capabilities
- No privilege escalation
- Resource limits defined
Continuous Monitoring
Runtime monitoring provides visibility into pod behavior:
- Process execution tracking
- File access monitoring
- Network connection detection
- System call auditing
Related Patterns
- Cluster Configuration
- Network Security
- IAM Configuration