Scaling AWS Fault Injection Service across your organization and accounts

The article discusses scaling AWS Fault Injection Service (FIS) across multiple accounts using centralized management strategies and robust IAM role configurations.

• Two multi-account experiment strategies are presented:
  • Centralized Management: One orchestrator account manages experiments
  • Decentralized Management: Each account runs its own experiments
• Key roles established:
• AWS-FIS-Experiment-Orchestrator: Creates and manages experiment templates
• AWS-FIS-Experiment-Target: Executes experiments in target accounts
• Benefits include:
• Enhanced security through role-based access control
• Centralized experiment management
• Improved organizational scalability
• Better compliance and audit capabilities

The approach enables systematic chaos engineering across AWS environments, helping organizations improve system resilience and reliability.