Skip to content
Katabench
Try free

Scale the Web Tier

Easy About 25 min Free with an account

Put a stateless app tier behind a load balancer and remove its single point of failure.

The brief

Requirement: A highly available API that keeps serving when it loses an instance. A single app process is a single point of failure, and one node cannot absorb growth or ride out a deploy without dropping requests.

The pattern: Put a load balancer in front of at least two stateless, interchangeable app instances. The balancer health-checks each one and reroutes around a failure, so losing an instance costs capacity rather than uptime.

Treat each app service node as one independently deployable, stateless instance. The topology should still serve requests when either app instance is removed. Session storage and retry behavior are operational concerns you will examine after the solve.

Drag components from the palette, wire them together, then Check your design.

Goals

  • Every App Service sits behind a Load Balancer.
  • The app tier runs at least two App Service nodes.
  • The Load Balancer fans traffic out to at least two App Service nodes.
  • A complete Client → Load Balancer → App Service → Database path exists, and every App Service reaches the Database.
  • Every client-to-database path passes through the Load Balancer.

The topology represents responsibilities and interaction semantics. You do not need to choose a specific load balancing algorithm, instance count, or deployment platform.

How it is graded

You build the design on a canvas, and Katabench checks the topology against the challenge's design rules.

System Design Studio →

Related guides

What it practices

Stateless app tier
Route clients through a trusted, horizontally scaled app tier behind a load balancer.

Related katas