What is Serverless Architecture? How is it structured and how does it work?

Question

IT Interview Practice · Accepted Answer

## What is serverless?

**Serverless** is a way to build and run applications in which a cloud provider manages the servers and provisions execution resources. The development team focuses on code, service configuration, and business logic instead of installing and maintaining servers.

Servers still exist, but the provider operates them. A common approach uses **functions**: code invoked by a request or an event. Beyond functions, providers also offer database and storage services that follow the serverless model.

## What makes up a function-based application?

- **Trigger:** An HTTP request, a schedule, or an event such as a newly uploaded image.
- **Function:** Code that performs the required task, such as creating a thumbnail.
- **Execution platform:** Prepares the runtime environment and adjusts the number of concurrent tasks within the service's limits.
- **Data storage:** A database or storage service keeps results durably. Function memory should not be relied on to preserve data between invocations.

## Example: Creating thumbnails automatically

When a user uploads an image to storage, an image-created event triggers a function. The function reads the original, creates a smaller version, and saves the result to a separate location. The application displays the thumbnail after processing finishes.

When many users upload images at once, the platform can run more functions concurrently within configured limits. The development team does not have to start additional servers for each burst of uploads.

If an event is delivered again, the function needs to handle it without creating duplicate results. For example, each original image has a unique identifier, and its thumbnail is always written to the same corresponding location. This is **idempotency**: repeating the same request preserves the same business outcome.

## Benefits

- **Less server administration:** The provider handles the infrastructure running the code, leaving the team more time to develop features.
- **Adaptable to variable workloads:** Processing capacity can increase or decrease with demand, such as a burst of image uploads followed by a quiet period.
- **Convenient for independent tasks:** A team can add an event-processing capability without setting up a separate server application.

## Limitations to understand

- **Startup can add latency:** When a new execution environment is needed, a request has to wait for initialization. This is a **cold start**, and it does not happen on every invocation.
- **Execution has limits:** Functions have service-specific limits on duration, resources, and concurrency. A large task may need to be divided or run elsewhere.
- **Failures still need attention:** The team must monitor failed tasks, configure retries, and handle duplicate requests; the provider does not automatically fix business logic errors.

## When should you use serverless?

Serverless functions suit image processing, notifications from other systems, scheduled jobs, and APIs with variable traffic. Consider the tradeoffs carefully when a task must run continuously for a long time or needs very consistent latency, because limits and pricing depend on the chosen service.

What is serverless?

What makes up a function-based application?

Example: Creating thumbnails automatically

Benefits

Limitations to understand

When should you use serverless?

Related questions