A company runs a production web application on Amazon EC2 that must stay on around the clock and has a stable, predictable load expected to continue for the next three years. The team wants the lowest hourly rate for this steady usage. Which pricing model best fits this need?
- AOn-Demand Instances, paying the standard per-second rate with no commitment for the running servers
- BSpot Instances, bidding on spare capacity to run the steady production workload at a deep discount
- CDedicated Hosts billed On-Demand, reserving a whole physical server for the steady production workload
- DReserved Instances, committing to a one-year or three-year term in exchange for a much lower hourly rate Correct
Why A is wrong: On-Demand suits short or unpredictable workloads, but for steady multi-year usage it costs more than a committed model, so it is the tempting default rather than the cheapest choice.
Why B is wrong: Spot offers large discounts but AWS can reclaim the capacity at short notice, which is unsuitable for an always-on production site that must not be interrupted.
Why C is wrong: Dedicated Hosts address licensing and isolation needs on physical servers, not cost savings, so paying that premium does not give the lowest rate for steady use.
Why D is correct: Reserved Instances reward a one-year or three-year commitment with a large discount over On-Demand, which matches a stable workload running continuously for years.