Serverless
Serverless runs code and services without managing servers, scaling automatically and billing per use, best suited to event-driven workloads.
Serverless is a cloud model where you deploy code or use managed services without provisioning or operating servers. The provider handles capacity, patching, scaling, and availability, and bills based on actual consumption rather than reserved capacity. Despite the name, servers still exist; they are simply abstracted away from the developer.
How It Works
The best-known serverless category is function-as-a-service, where short-lived functions run in response to events such as HTTP requests, queue messages, or file uploads. The platform starts an instance when an event arrives, runs the function, and may keep it warm or shut it down. Scaling is automatic, from zero to many concurrent executions. Serverless also covers managed databases, queues, and storage that scale and bill per use. Because instances are ephemeral and stateless, applications externalize state to databases, object storage, or caches.
Why It Matters
Serverless removes most operational burden and aligns cost tightly with usage, which is attractive for spiky, event-driven, or low-traffic workloads. Teams ship faster because there is no server fleet to manage. The trade-offs include cold-start latency when idle functions must initialize, execution time and resource limits, harder local testing, and the risk of vendor lock-in since serverless platforms differ. High-volume steady workloads can sometimes be cheaper on always-on containers or reserved instances, so teams model cost before committing.
Related Terms
Serverless is most often implemented as function-as-a-service and is closely tied to cold starts, autoscaling, and managed services.