No content

Building reactive applications with Node.js and Red Hat JBoss Data Grid Galder Zamarreño Arrizabalaga Divya Mehra 

Galder Zamarreño Arrizabalaga • Senior Software Engineer @ Red Hat • Infinispan / JBoss Data Grid co-founder • JSR-107 expert group member • Keen interest in functional programming 

Divya Mehra • Principal Product Manager @ Red Hat • Responsible for Red Hat JBoss Data Grid roadmap 

Agenda • Introduction to Red Hat JBoss Data Grid • Building a reactive application with Node.js and Red Hat JBoss Data Grid 

Introduction to Red Hat JBoss Data Grid 

Red Hat JBoss Data Grid 

Why in-memory data management? • Large volume, velocity, and variability of raw data • Enterprises that can extract meaningful, actionable information faster have competitive advantage • Increase revenues • Reduce risks 

E-commerce website Challenge: Maintain page load times, while computing relevant purchase recommendations from data from different sources: • Mobile device → location • Social graph →what friends are buying or recommending • 3rd party sellers or partner sites -> current promotions • User’s purchase or click-through history -> behavioral targeting Relevant recommendation can increase conversion by 25% However, each 1-second increase in page load time → up 10% drop in overall revenue Can your data platform scale out to large volumes of often-changing data without affecting application performance? 

SaaS Application Provider Challenge: Users of SaaS application should be unaffected by spikes in demand, node failure, and system upgrades of the data platform. Loss of availability and performance will affect SaaS app provider’s SLA to its customers • Revenue loss • Reputation impact Does your data platform support high availability, elasticity, and rolling upgrades, necessary for a cloud-native architecture? 

Telecom service provider Challenge: Real-time billing and usage information for thousands of concurrent users • A customer sends a text, uses data, or makes a call • Transaction is recorded in the data platform, and the customer’s total usage and bill is updated immediately triggered by creation of this new entry. Can your business logic react to changes in the data in real-time and at scale? 

JBoss Data Grid - 4-IN-1 package 

Leader in Forrester Wave™: In-Memory Data Grids, Q3 2015 Ahead in both evaluation dimensions vs. open source competitors 1.Current offering 2.Strategy and vision Download free from: https://engage.redhat.com/ forrester-data-grid- s-201509240128 

Reactive apps with Node.js and JBoss Data Grid 7 

Architecture HTTP TCP/IP 

Elm • Functional language for Web • Compiles to Javascript • No runtime errors in practice • Statically typed, friendly errors • Help build well-architected code • Easy to learn! Easy to use! 

Event Manager Microservice • Node.js middleware • Based on Express.js • Listens for HTTP on port 3000 • Uses JDG 7 Node.js client • Requires Node 0.10 or higher 

JDG Persistence • Runs a JDG 7 server domain • Formed of 3 nodes • Each element stored in 2 nodes • Redundancy via partial replication • Redundancy for failover 

Event Iteration (1) GET /events 

Event Iteration (2) GET /events iterate event 1 event 2 … 

Event Iteration (3) GET /events iterate event 1 event 2 … JSON event[] 

Demo Event Iteration 

Event Insert (1) POST /events 

Event Insert (2) POST /events putIfAbsent 

Event Insert (3) POST /events putIfAbsent true/false created event 

Event Insert (4) POST /events putIfAbsent true/false created event JSON true/false WS: JSON event WS: JSON event 

Demo Event Insert 

Event Search (1) GET /search ?q=query 

Event Search (2) execute GET /search ?q=query 

Event Search (3) execute results JSON results GET /search ?q=query 

Demo Event Search 

github.com/galderz/infinispan-events/tree/june16 

Credits Window by Oleg Frolov from the Noun Project Server Error by Montu Yadav from the Noun Project Databases by Oliviu Stoian from the Noun Project