Explore full-stack Flux architecture in this conference talk from React.js Conf 2015. Dive into the challenges of current backend systems and discover how Flux can address these issues. Learn about Facebook chat implementation, the concept of shared mutable state, and scaling actions and queries. Examine the backend perspective, understanding applications as derived data from action history. Investigate techniques for efficient action processing, data store management, and choosing optimal data stores for specific queries. Analyze communication between stores, data enrichment strategies, and the implications of consistency models. Gain insights into stream processing frameworks and the concept of turning databases inside out for improved application architecture.
Full-Stack Flux Architecture in React.js - React.js Conf 2015
Overview
Syllabus
Intro
Interesting abstraction
Solvers are unpredictable
Full-stack Flux
The current state of the backend
How Flux can save us
A review of Flux
Facebook chat
The old way
A new question
A new action
Alternative: observe a data model
Shared mutable state is the root of all evil.
Flux helps you scale actions and questions
What about the backend?
Your app == derived data ...using the history of every action that ever happened
Process every action for every query request.
Can we make it fast enough?
An example: photos
Lots of advantages Manage the trade-atts of your data store granularly
Choose the best data store for your question.
Closing the loop
Things that don't work
Communication between stores waitfor() doesn't work
Option 2: Enrich data Consume action, emit new action with additional data.
Strong consistency Only individual answers to questions may be consistent
Strong consistency Lack of strong consistency is a more accurate worldview
Read about stream processing frameworks.
Turning the Database Inside Out
Taught by
Meta Developers
