REFACTORING OF MONOLITH TO MICROSERVICES ****

Udagram is a simple cloud application developed alongside the Udacity Cloud Engineering Nanodegree. It allows users to register and log into a web client, post photos to the feed, and process photos using an image filtering microservice.

The project is split into two parts:

1. Frontend - Angular web application built with Ionic Framework
2. Backend RESTful API - Node-Express application

Container and Microservices

Divide an application into microservices

Refactoring the Udagram Backend API

• Backend-api here
  • Backend-feed here - /feed
  • Backend-users here - /users

The Backend API was separeated into two microservices API as independent services.

Build and run a container image using Docker

The project includes Dockerfiles to successfully create Docker images for /feed, /user backends, project frontend, and reverse proxy.

Dockerfiles

• Backend Dockerfile
  • Feed here
  • Users here
• Reverse-Proxy Dockerfile here
• Frontend Dockerfile here
• Base Image here

DockerHub Images

• Backend Images

Fig 1.0: NodeJs Base Image

Fig 1.1: Backend Users Docker Image

Fig 1.2: Backend Feed Docker Image

Reverse-Proxy

Fig 1.3: Reverse Proxy Docker Image

Fig 1.4: Frontend Docker Image

Independent Releases and Deployments

• Travis Config File

With Travis CI, the independent releases and deployment was automated following DevOps and Security best practices.

The configuration file .travis.yml can be found in the Monolith-Microservices source repo.

Fig 1.5: Travis Configuration File

• Travis Interface - Deployment Status

Fig 1.6: Travis Interface

Service Orchestration with Kubernetes

• Create ConfigMap

Config file or Config.yaml file

POSTGRES_HOST: host_address
POSTGRES_DB: database_name
AWS_BUCKET: bucket_name
AWS_REGION: region
AWS_PROFILE: profile_name
JWT_SECRET: jwt_token
URL: http:

kubectl create -f confimap <name.yaml>

or 

kubectl create configmap <configmap_name> --from-file config

• Create Secret

Encode the Data

base64 login_credentials

Create the secret.yaml file

kind: Secret
metadata:
  name: env-secret
type: Opaque
data:
  POSTGRES_USERNAME: <your_base64_encoded_data>
  POSTGRES_PASSWORD: <your_base64_encoded_data>

• Deployed Microservices using a Kubernetes cluster on AWS

  • kubectl get pods

  

  Fig 2.1: Deployed Resources

  
  
  • kubectl describe services

  

  

  

  Fig 2.2: Sensitive Data Not Expose

  
  
  • Reverse proxy to direct requests to the appropriate backend

  

  Fig 2.3: Reverse Proxy

  
  

  • Scaling and Self-healing for each service

    • Replicas Defined in Deployment.yaml

    

    Fig 2.4: Deployment File with 2 Replicas

    kubectl apply -f deployment.yaml

    kubectl get pods

    

    Fig 2.5: Deployment Status

    
    

    Note: The next steps depends on Metric Server

    Installing metric server can be done in one of the following

    1. Kubernetes Metric Server

    kubectl apply -f https://github.com/kubernetes-sigs/metrics-server/releases/latest/download/components.yaml

    2. Helm charts

    Note: Previous Metric Server on helm is deprecated, use the prometheus adaptor instead.

    helm install --name my-release stable/prometheus-adapter

  • Autoscaling with CPU Metrics

    kubectl describe hpa backend-api

    

    Fig 2.6: Autoscaling with CPU Metrics

    
    

    Debugging, Monitoring, and Logging

    

    Fig 2.7: Debugging DB Connection Request

    Problems

    Solutions