Test Scenarios

TODO: Move this to separate pages with deeper documentation of each item:

Deployment, Elastic/Web Scale

• Simple deploy a fleet of each tier of an application
• Automatically (or manually?) deploy same fleet across multiple fault domains (availability zones)
• Automatically (or manually?) deploy parts of feel on single optimized domain (placement group)
• Manually elastically scale a fleet of services
• Load, SLA, or response time scaling of deployment
• Force re-balancing of zones of deployment

High availability, Automatic recovery, Disaster Recovery

• Investigate following failures
• Instance failure/recovery
• Entire fault domain failure and HA
• Network partition between fault domains
• Sick instances
• still running, but erroring
• still running, not erroring on all request, but app centric health check failing
• Show durable backup of data volumes and recovery of data
• Durable storage reclaims for failed containers (database that wasn't memory replicated)

Devops (Continuous Delivery)

• Demonstrate, without downtime, canary deployments
• Demonstrate, without downtime, red/black deployments
• Demonstrate, without downtime, rolling upgrade deployments

Operational Visibility

• Single instance level metrics
• Fleet level aggregated metrics
• Micro-service interaction metrics
• Remote logging and fleet level analysis

Misc

• Show integration with network security (security groups, firewalls, etc.)
• Show integration with other network facilities (front end public internet facing load balancers)
• Isolation testing (network, disk, compute, etc.)