Visualizers
Learn by moving things
Interactive explainers for the ideas behind retrieval, embeddings, and agents. A formula tells you what is true; dragging a vector shows you why. Each one also appears inside the article it belongs to, and runs entirely in your browser, so drag, poke, and break it.
Round-robin
From: How a Load Balancer Spreads the Load →The simplest rule: deal each request to the next server in turn, around and around. Even when servers are identical, but blind to how busy any one of them is.
Round-robin: each server gets the next request in turn, like dealing cards. Even and simple, but it ignores how busy each one is.
Weighted round-robin
From: How a Load Balancer Spreads the Load →Round-robin in proportion to a weight, so a bigger server gets a bigger share. Also how canary releases work: give the new version weight 1 and the old one weight 9.
Weighted round-robin: bigger servers get a bigger share of the rotation. Server 3 has weight 3, so it is dealt three times as often as server 1.
Least connections
From: How a Load Balancer Spreads the Load →A dynamic rule: send each request to the server handling the fewest connections right now. Great for long-lived, uneven work like database sessions or WebSockets.
Least connections: send the request to whichever server is handling the fewest right now. Good when connections are long-lived and uneven.