A company operates a production server with a nearly constant workload running 24 hours a day, 365 days a year. The company wants to receive a significant discount by committing to 1 or 3 years of usage. Which purchasing option is MOST suitable?
A question about selecting the MOST suitable long-term discount purchasing option for an always-on production server.
Spot Instances
This is incorrect. Spot Instances are a purchasing option that offers up to approximately 90% discount on AWS spare capacity, but can be interrupted by AWS with a 2-minute notification. Because of the interruption risk, they are not suitable for a production server that must run continuously 24/7.
This option does not meet the requirement for a stable discount through a long-term commitment.
On-Demand Instances only
This is incorrect. On-Demand is flexible but offers no discount, making it costly for always-on workloads.
This does not meet the requirement of receiving a discount by committing to long-term usage.
Savings Plans or Reserved Instances (RI)
This is correct. Savings Plans and Reserved Instances (RI) are purchasing options that provide a significant discount over On-Demand pricing in exchange for committing to 1 or 3 years of usage. For workloads that run stably for long periods, these long-term commitments provide the MOST cost-effective option.
As a key difference: Reserved Instances commit to a specific instance type and region (and can reserve capacity), while Savings Plans commit to an hourly usage amount ($/hr) and apply flexibly across instance families, regions, Lambda, and Fargate.
Dedicated Hosts
This is incorrect. Dedicated Hosts are a purchasing option that dedicates a physical server and are intended for license compliance or regulatory requirements.
They are not primarily chosen for the purpose of receiving a long-term commitment discount.
'Stable workload,' '1–3 year commitment,' and 'significant discount' all point to Savings Plans / Reserved Instances. For the lowest cost with interruption tolerance, use Spot. For short-term or unpredictable usage, use On-Demand.