Load Balancer
System Analysis
Normal Behavior
Distributes traffic evenly across healthy nodes to ensure high availability.
Failure Behavior
Routes 100% of the traffic to the single node that is currently restarting.
Business Consequence
Takes down the entire cluster, one node at a time, in a perfectly orchestrated rolling blackout.
Visual Manifestation
"A frantic traffic cop directing all 18-wheelers onto a collapsing wooden bridge."
Satirical Behavior
"The incredibly stressed middle manager of your infrastructure whose only job is to hand angry customers to whichever server looks the least dead."
Known Aliases
Technical Terminology
Failure Indicators
System Architecture (Graph)
FAQ
How does it normally behave?
Distributes traffic evenly across healthy nodes to ensure high availability.
How does it fail?
Routes 100% of the traffic to the single node that is currently restarting.
What is the business consequence?
Takes down the entire cluster, one node at a time, in a perfectly orchestrated rolling blackout.
What is the difference between Layer 4 and Layer 7 load balancing and when should each be used?
Layer 4 (L4) load balancers operate at the transport layer (TCP/UDP), routing packets based on IP addresses and ports without inspecting application payloads; they offer ultra-high throughput and microsecond latency, ideal for gaming or raw databases. Layer 7 (L7) load balancers inspect application-layer data (HTTP headers, cookies, URLs), enabling intelligent path-based routing, TLS termination, and WAF filtering at the cost of higher CPU overhead.
How does health check flapping create a cascading failure across backend server pools?
When a backend pool experiences a temporary traffic spike, servers slow down their response times. If the load balancer's health check timeout is too strict, it incorrectly marks these slow nodes as dead and removes them from the pool. The remaining healthy nodes must absorb the entire traffic load, causing them to slow down and fail health checks as well, resulting in the rapid eviction of all nodes in a cascading total outage.
Explore the system
AI Summary
Load Balancer is a NETWORKING system in TinyCTO.tv. Accepts client connections at a single virtual IP (VIP), evaluates load balancing algorithms (Round Robin, Least Connections, IP Hash), runs periodic health checks against registered backend targets, routes traffic exclusively to healthy nodes, and manages connection keep-alive pools.