High Availability and Scalability in the Cloud

Scenario: short-lived traffic spike

Scale up for the spike, then scale down to control cost. Category: scalability. Scale up to stay responsive. Scale down after the event. Avoid paying for idle peak capacity.