Question

A media company needs to process user-uploaded images by generating thumbnails. Uploads are highly unpredictable, ranging from zero to thousands per minute during viral events. The team wants to avoid managing servers and pay only for actual processing time, with automatic scaling to handle sudden spikes. Which architecture best meets these requirements?

Accepted Answer

Deploy a function-as-a-service (FaaS) that triggers on object-storage upload events to process each image

Answer

Provision a fixed-size autoscaling group of virtual machines behind a load balancer polling a queue

Answer

Run a containerized processing service on a dedicated Kubernetes cluster sized for peak load

Answer

Use a single large compute-optimized instance with a scheduled batch job to process uploads hourly