Load Balancer Pool Distribution Analyzer

Also available in:

Load Balancer Pool Distribution Analyzer

Simulate how different load balancing algorithms distribute traffic across pool members. Model failure scenarios and compare redistribution efficiency.

Load Balancing Algorithm:

Virtual Nodes (Consistent Hash):

Traffic Volume

Sessions to simulate: 1000

Pool Members

Member IP Address:

Port:

Weight (1-100):

Failure Scenario Simulation

Target Member:

How it works: Client IPs are sampled from a /16 subnet (10.0.0.0/16). For source-based hashing, each unique client IP is deterministically mapped to a member. Round robin ignores client IP and distributes sequentially. Weighted uses proportional random assignment. Consistent hashing places both clients and members on a virtual ring.

Pool Members

ID	IP:Port	Weight	Status	Actions

Distribution Results

Session Distribution per Member

Failure Redistribution Analysis

Consistent Hash Ring

Reading the ring: Each member is placed at multiple points around the ring (virtual nodes). A client IP is hashed to a position on the ring, then the request is sent to the nearest member clockwise. Adding or removing a member only affects the arc of the ring it owned — minimizing session disruption.

Algorithm Reference

Round Robin

Sessions are assigned sequentially: member 1, 2, 3, ..., N, 1, 2, ... Each member receives exactly 1/N of all requests. Ignores server capacity and client affinity. Simple and predictable but fails hard on heterogeneous hardware.

Source IP Hash

A hash of the client IP modulo the member count selects the backend. The same client always reaches the same member — useful for stateful applications. Adding or removing a member reshuffles all client mappings (N change to (N-1)/N of all sessions).

Weighted

Each member receives a share proportional to its weight relative to the total weight pool. A member with weight=4 gets 4x the sessions of a weight=1 member. Used to model heterogeneous backend capacity (e.g., a VM vs a bare-metal server).

Consistent Hashing

Members and clients are both mapped onto a circular ring via hashing. Each client goes to the nearest member clockwise on the ring. Virtual nodes (replicas per member) improve distribution uniformity. When a member fails, only its sessions move to the next member on the ring — 1/N sessions disturbed vs. 100% for modulo hashing.