AWS Reliability Pillar

Design Principles

• Automatically recover from failure – use KPIs to trigger automatic system recovery
• Test automatic recovery – validate recovery procedures
• Scale horizontally to increase aggregate workload availability – use autoscaling
• Stop guessing capacity – monitor demand and utilization to trigger scaling in or out
• Manage change in automation – automate all changes to infrastructure for reliable recovery

Best Practices

• Foundations – consider service quotas and network capacity
• Workload architecture – design failure prevention and failure mitigation
• Change management – design for changes in demand and capacity with monitoring and triggering in response to KPI changes
• Failure Management – failure detection and automatic repair, backup and recovery, DR planning and testing

Services

• AutoScaling
• AWS Backup
• AWS Cloudwatch