Load Balancing
Definition
Load balancing distributes incoming requests across multiple instances or backends.
Methods can use round robin, least connections, weighted, or health-based strategies. Load balancing supports scaling and availability but does not replace stateful architecture.
Simply explained
How to distribute incoming work across multiple available switches.
Why it matters in practice
Relevant for highly available websites, APIs, and inference services.
Example from business
An API gateway distributes requests only to healthy backend instances.
Limits and pitfalls
- Missing health checks; uncontrolled session affinity; misunderstanding load balancers as database scaling.
Origin & history
Load balancing is a fundamental principle of distributed systems and modern cloud platforms.
Use in mid-sized companies
- Possible application in SMEs: Controlled distribution of multiple web, worker, or model instances.
Frequently asked questions
Is load balancing the same as autoscaling?
No. Are Health Checks needed? Typically yes.
Sources
- Service LoadBalancer · Kubernetes, Service LoadBalancer
- Ergänzender Primärbeleg · Ergänzender Primärbeleg
Source verification: 10./12.10.2026.
Editorially reviewed on: