AWS Reserved Instances (Amazon Reserved Instances) are virtual servers that run in Amazon Web Services' Elastic Compute Cloud (EC2) and Relational Database Service (RDS). The companies purchase the instances at contract prices, plus hourly rates. Instances are available in varied levels of compute power.
Amazon EC2 offers three pricing options for instances: On-Demand Instances, Reserved Instances and Spot Instances. IT administrators can purchase Reserved Instances directly from Amazon Web Services or from other admins on the AWS Reserved Instance Marketplace. Reserved Instances purchased in the Marketplace can provide cost savings for computing power for shorter remaining terms than new instances. Users can opt for no upfront payment, partial payment or to pay in full when purchasing Reserved Instances from AWS. Contract terms last one or three years.
To create a Reserved Instance, companies choose one of Amazon's various instance types and an associated operating system (OS), term, tenancy specification, region and availability zone. Instances can be exchanged within the instance family or moved between different availability zones within the region.
AWS Reserved Instances in RDS are available in three usage tiers in a contract period: light, medium and heavy. Light Reserved Instances are used less than 39% of the contract period time, while medium instances are used less than 75% of the time. Heavy Reserved Instances are in use for more than 75% of the time. IT teams can shut down light and medium instances to reduce hourly rates, but heavy instances are always billed hourly.
AWS Reserved Instances in both EC2 and RDS are available for 750 hours of free service per month; EC2 requires free-tier users to use only micro-size instances -- low-cost instances with low to moderate networking performance -- while RDS users are restricted to 20 GB of storage and 20 GB for backups in the free tier.