At the core of the Edge Gateway service lives a request queue that remained relatively unchanged through the majority its existence. It was stable, performant and resourceful, but ultimately ill-suited to the ultimate goal; to distribute work to devices prioritised by their absolute real-time relative performance grade. — Centralised load balancing As those of you who have ever had to horizontally scale a services deployment with a typical cloud hosting provider will know, the mechanics are fairly simple. You create a replicable ‘image’, set some limits, and sit back and watch y̵o̵u̵r̵ ̵b̵a̵n̵k̵ ̵b̵a̵l̵a̵n̵c̵e̵ as server-after-server spawn in the wake of…