Case Study: Twitter powers real-time tweet search updates with Cloud Foundry

Twitter runs a Spring worker polling every 5 seconds on Cloud Foundry

Twitter utilized Cloud Foundry's platform to deploy a standalone Java worker application built with Spring Integration. The challenge was to create a background process that could efficiently poll Twitter for specific content and publish it for consumption by a separate web front-end, requiring a robust and scalable cloud infrastructure.

The solution implemented by Cloud Foundry involved running a Spring-powered worker that queried Twitter every five seconds for tweets containing the word "cloud" and published the results to a shared RabbitMQ service. A Node.js web application, also deployed on Cloud Foundry and bound to the same service, then consumed these tweets and displayed them in a dynamically updating web page. This architecture, facilitated by Cloud Foundry, successfully decoupled the worker and web processes, allowing them to scale independently and resulting in a fully functional, real-time application.

Twitter