Streaming and Push Patterns¶

When data arrives continuously or events occur asynchronously, pushing data from producer to consumer is more efficient than repeated polling; this inverts the control flow from pull-based request-response.

Tension Addressed¶

Simplicity vs. Performance; Statelessness vs. Efficiency

Recognizable As¶

WebSockets, server-sent events, pub/sub, observer pattern, reactive streams, callbacks, event loops

Canonical Sources¶

Hohpe & Woolf, "Enterprise Integration Patterns" (2003); Reactive Manifesto (2014)

Transfers To¶

Event-driven architecture, real-time systems, UI frameworks (reactive updates), message brokers, log tailing

Sub-concepts: - Long polling as hybrid: Simulating push over request-response by delaying the response - Multiplexing streams: Multiple logical streams over one physical connection