Introduction to Reserving Capacity
To provide greater pricing flexibility, many AWS services offer reserved capacity plans. When you commit to reserved capacity, you pay for the resource regardless of actual usage over the life of the reservation.
Compared to the on-demand rates that only charge for what you use, reserved capacity offers significant cost savings. If your service usage is predictable and will extend long-term, alternative billing models will drive down costs. Reservations are typically made one or three years in advance.
Some reserved services, such as EC2 Reserved Instances, can guarantee capacity in an availability zone. This protects large and location-dependent environments against regional spikes in usage and reduced availability during outages. Priority over other types of pricing plans means organisations always have access to the resources they need to continue critical business operations.
What Services Can I Reserve?
Amazon continues to add reserved pricing models for services where customers demand flexibility. In particular, look around your environment for:
EC2 instances. You can reserve compute capacity for EC2 instances with AWS Reserved Instances for a saving of up to 75% over on-demand billing. As opposed to other reserved services, reserved EC2 instances don’t have to be for one or three years; you can customise with on-demand capacity reservations. However, reservations are specific to instance type, platform and optionally the availability zone.
Amazon Relational Database Services (RDS). Capacity reservation is available for databases running Aurora, MySQL, MariaDB, PostgreSQL and SQL server engines, for a cost reduction of up to 69%. If your data is persistent and needs to be available at all times, databases are the perfect service to reserve.
Amazon RedShift data warehouses offer reserved capacity pricing.
Amazon CloudFront configurations, if your environment sends terabytes of data through the CDN.
How to Reserve Services
Reserved capacity plans can be purchased directly from the AWS Management Console. Some services, such as Amazon CloudFront, require that you first contact AWS to organise a plan.
Once you make a reservation, the cost savings are applied to your current eligible services. Payment is either with all upfront, partially upfront or no upfront, with slight discounts for upfront payments.
Although reserving capacity is an excellent cost optimisation strategy, identifying appropriate services to reserve is a difficult task. Organisations are billed for the reserved service regardless of their utilisation. This makes mistakes costly and might prevent you from making optimal changes to your environment.
First Steps
A prudent approach analyses the historical service-by-service usage for opportunities to reduce costs. Begin by implementing a Cost Management Platform (CMP) to start gathering valuable insights, data records and prompt your investigations.
Consider the option of reserved capacity holistically. In addition to other pricing models, such as spot pricing or savings plans, reserving the service needs to be compared against other cost-saving opportunities, including consolidating, right-sizing or closing them.
This investigation will lead your business towards building a cost-optimised environment that meets your requirements whilst maximising project return on investment.