Load balancers are discussed a lot more often these days than they used to be, say, a decade or two ago. They used to be somewhat arcane tools that only some network planners or architects really had to worry about during edge-case planning; now, they are absolutely mainstream, and even developers and app architects need to understand what kind of load balancer to choose, and why.
Why have load balancers become such a conversation starter these days? The answer lies in two important features of compute on the cloud—ephemeral external IP addresses and autoscaling of backends:
In the cloud world, load-balancer devices are an essential static entry point for apps. They have a static IP address that clients can be sure will remain unchanged, and accept incoming client requests and distribute them to a variable set of backend instances. The size...