Microservices vs Monolith

MicroService

• Take every app function and put it in its own service
• Runs in its own container
• Communicate via APIs

But if each microservice is implemented in different languages/environment, how do we deploy? Deployment: Use Docker ! Put each microservice into containers

Pros

• Flexibility (each microservice can be built in different languages/technologies)
• Less risk in change
• Independent scaling
• Faster release cycles

Monolith

• Server-side systme based on single application

Pros

• Good for small team
• Less complex
• Less duplication
• Run fast

Cons

• Highly dependent (single point of failure)
• Language/Framework
• Growth
• Scaling issue
• Complex deployment

Spring Boot

Goal

• Enable quickly built applications
• Provide common non-functional features

Not

• Spring boot does NOT generate code
• Spring boot is NOT an application server or a web server

Features

Quick starter projects with auto configuration

• Web (Login, spring mvc)
• JPA (DB)

start.spring.io starter apps/dependencies

• Spring Web
  • @RestController
    • @GetMapping handles REST API
  • MVC
• Spring Boot Actuator
  • Monitor and manage application
• Spring Boot DevTools
  • Provide fast app restarts, livereload

Spring vs Spring Boot vs Spring MVC

Spring Framework

• Dependency Injection: Takes care of defining beans, dependencies, how to bind components

Spring MVC

• Web applications

Spring Boot

• Autoconfig

// 143

Questions: What is spring boot/cloud? Distirbuted system regarding microservice load balancer what does it do? what is maven

$$$$$$$$$$ Draw a overall blueprint of each model Why use these techs $$$$$$$$$$

Keyword

• Distributed system
• Maven
• Java Spring Boot? Dependencies:
  • Spring Boot Devtools
  • Spring Boot Actuator
  • Spring Web(REST, MVC)
  • Config Client (Connect to spring cloud config)
• REST
• H2 DB, SQL with JPA (Currency from / to, exchange rate)
• Build routes with Eureka naming server and redirect all requests thru Spring cloud gateway (Assigen different ports to different server)
• Circuit Breaker framwork: Resilience4j (fallback methods if ms is down)
• Zipkin as a distributed tracing server (as a container) connected to all ms ( to trace requests accross ms) // 174
  • What if distributed tracing server is down? Use Rabbit MQ
• Google Cloud Kubernetes