Level 2 · Systems
Load balancing and scaling
Traffic triples at 09:00 every weekday and the site falls over at 09:02. Adding a bigger instance fixes it until the day traffic quadruples. Horizontal scaling is a different shape of answer, and it forces your application to become stateless.
Chapters
- How load balancers route incoming trafficClients need a single entry point that survives server failures. A load balancer decouples incoming requests from individual backend instances to keep systems online.
- Why scaling out takes longer than expectedAdding servers automatically is not instantaneous, and understanding the delay between a metric spike and a running application is key to preventing system outages.
- What a health check actually provesA health check determines whether a server is capable of processing work, but poorly designed checks can easily trigger self-inflicted outages.