Reactive: Vert.x & RSocket

RSocket's long-lived stateful connections clash with stateless scale-to-zero functions, so keep the persistent connection on an always-on edge and let serverless handle short, stateless units of work.

RSocket backpressure is demand-based at the message level via request(n), while HTTP/2 flow control is byte-window-based at the transport level, so RSocket matches producer rate to what the application can actually process.

Handle per-stream errors with onErrorResume and retryWhen, detect dead peers with keep-alive, and survive drops with resumption tokens; without resumption, streams error out and the client reconnects.

Scale RSocket with connection-sticky or client-side load balancing, autoscale on connection metrics, enable resumable sessions for pod churn, and tune keep-alive to tolerate rolling updates.

RSocket recovers with request-level retryWhen backoff for transient faults, resumable sessions for dropped connections, and circuit breakers to fail fast and protect a struggling peer.

Vert.x serves massive concurrency with few threads by never blocking the event loop and passing messages asynchronously over an event bus.

Schedule the trigger on the event loop with setPeriodic or a cron timer, but always run the heavy batch work on a worker verticle or executeBlocking.

RSocket is a binary reactive protocol with four interaction models and built-in application-level backpressure, where either peer can initiate and streaming is first-class.