High Availability and Fault Tolerance
Auto Scaling alone does not guarantee resilience. It must be combined with multi-AZ deployment.
Multi-AZ Strategy
An Auto Scaling Group can span multiple Availability Zones.
This ensures:
- Traffic distribution across zones
- Automatic replacement if one AZ fails
- Higher availability SLA
Architectural principle: Never deploy an ASG in a single Availability Zone for production.
Health Checks
Two health check types:
- EC2 health checks
- Load Balancer health checks
In production, always use load balancer health checks.
Why?
Because infrastructure-level health does not guarantee application health.
Self-Healing Architecture
Auto Scaling continuously:
- Detects unhealthy instances
- Terminates them
- Replaces them
This creates a self-healing system.
In distributed systems, failure is expected — not exceptional.
Auto Scaling operationalizes this assumption.
In this section, I learned:
0 of 4 completed