A team is deciding which of its Amazon EC2 workloads it can safely move onto Spot Instances to capture the deepest possible discount, accepting that AWS can reclaim the capacity at short notice. Which TWO workloads are the best fit for Spot Instances? (Select TWO.)
- AA nightly batch rendering job that writes checkpoints and can resume cleanly after any interruption. Correct
- BA primary production database holding customer orders that must stay available every hour of the day.
- CA stateless fleet of automated test runners whose jobs can be retried freely if an instance is reclaimed. Correct
- DA licensed application that a vendor rule requires to run alone on a physical server reserved for one tenant.
- EA steady internal web portal kept running around the clock for three predictable years at a fixed size.
Why A is correct: Fault-tolerant batch work that checkpoints and resumes is the classic Spot fit because reclaimed capacity costs no lost progress.
Why B is wrong: An always-on production database cannot tolerate sudden reclamation, so Spot is unsafe and a steady-baseline model fits instead.
Why C is correct: Stateless, retry-safe test runners lose nothing when an instance is reclaimed, making them an ideal interruptible Spot workload.
Why D is wrong: Per-socket licensing on isolated hardware points to Dedicated Hosts, a need Spot Instances were never designed to meet.
Why E is wrong: A steady three-year baseline suits a committed Reserved Instance or Savings Plan, not the interruptible Spot model.