Guide to setup AWS auto scaleble server based on Dockered Spring Boot REST API server.

In progress

Create application

Build server

 gradle build

Build docker image

docker build -t  sample_api_img . 

Run docker container

docker run --rm -p 8080:8080 sample_api_img

Push to docker-hub

docker login
docker tag sample_api_img $DOCKER_USER/sample_api:dev
docker push $DOCKER_USER/sample_api:dev

Run docker from docker-hub

docker run --rm -p 8080:8080 $DOCKER_USER/sample_api:dev

AWS setup

Here we will configure AWS to scale out server automaticaly when load is increassed and scale in when load in descresed. In order to achive it we will create instance image which will be used and base image to setup new instances. Then setup loadbalancer to balance traffic between all instances.

Create security groups services>EC2>Security Group

Create two groups one to controll load balancer traffic and another to controll web tire traffic.

• Create group for Load Balancer
  • Click Create Security Group and set name for load balancer group
  • Add Rule for inbound traffic
  • Add Rule for outbound traffic
  • Click Create.
• Create group for Web Tire
  • Click Click Create Security Group and set name for web-tire
  • Add Rule for inbound traffic (http)
  • Add Rule for ssh port 22
  • Input load balance security Group Name in field Source to restrict traffic for balancer only

Create key pair

Private public key pair is used to access AWS linux based instances. And created under Services > EC2 > Key Pairs and click "Create Key Pair".

Create EC2 instance

First lets create instance which will be be used to create image. Then this image is used to created other instances. Navigate to Services > EC2 > Instances and click Launch Instance.

• Select server. For this example we are going to use Ubuntu Server and instabce type t2.micro which is available in trial account.
• Click Configure Instance Details
• Select availability zone as subnet aattribute (e.g. zone a)
• Left all other parameters with default values
• Clicke "Next" here we left default paramaters.
• Click "Next" and add tag to mark this instance (e.g. 'sample-ami')
• Click "Next" and choose sequrity group created earlier.
• Click Review and Launch
• Select creates earlier key pair and Launch instance
• Open instance and give name
• ssh to created instance

$ sudo ssh -i "./sample-key.pem" ubuntu@instance_public_dns

• Install Docker by commanda below or follw instraction on https://www.docker.com/

curl -fsSL https://download.docker.com/linux/ubuntu/gpg | sudo apt-key add -
sudo add-apt-repository "deb [arch=amd64] https://download.docker.com/linux/ubuntu $(lsb_release -cs) stable"
sudo apt-get update
apt-cache policy docker-ce
sudo apt-get install -y docker-ce

• Login to your docker account

docker login

After that docker file with credentials will be stored in ~/.docker/config.json.

Add application start commands into startup script

sudo vim /etc/rc.local

• Add run commands just befor exit 0

su - ubuntu
export DOCKER_CONFIG=/home/ubuntu/.docker/
export SAMPLE_APP_ENV=dev
export DOCKER_CONFIG=/home/ubuntu/.docker/
docker pull maksymenko/sample_api:$SAMPLE_APP_ENV
docker run -d --rm -p 8080:8080 $DOCKER_USER/sample_api:$SAMPLE_APP_ENV

Just left this file here so that all instances will have this file shen new instance created to scake uot application.

Create AMI

Navigate to Services > EC2 > Instances select instance created above.

• Click Action and select Image>Create Image
• Give name of image (e.g.: sample-image)
• Click Create Image
• Go yo AMI section to see that image creation is in progress, it take some time for new AMI to show up.

Create EC2 instance based on created AMI image

• Go to AMI section \
• Choose AMI created above
• Click Launch
• Click Configure Instance Details
• Select availability zone as subnet aattribute (e.g. zone a)
• Left all other parameters with default values
• Go to Advanced Details section and specify bash command which will be executes when instance created. Here we are going to specify custom bash commands tspecific for instance.

#!/bin/bash

• Click Next
• Choose Security Group as sample-web-tire
• Click Launch
• Use log file to debug script behavir
  • /var/log/cloud-init-output.log
  • /var/log/syslog

Check Api in browser

• Find public DNS of created instance and point browser to port 8080 http://{instance_public_dns}:8080

Setup Load balancer

• Navigate to Services > EC2 > Load Balancers and click Create Load Balancer
• Choose Application Load Balancer and click Create
• Give name (e.g. sample-app-loadbalancer)
• Set port 8080 for http as our application uses port 8080
• Add https is recommended for production usage (here we can skip it)
• Select availability zones, lets choose at least to zones a and zone b for system reliability.
• Define tags to mark balancer.
• Click Next and agail Next to choose Security Group created above.
• Selected existing security group created to controll loadbalancer traffic.
• Click Next to configure routing and set port 8080
• Create new target group to which load balancer will send traffic.
• Give name to group and click Next to register targets instances
• Select instances and add to registered on port 8080.
• Click Next to review and Create

Setup Autoscaling

In order to setup autoscaling we have to define two components "Launch Configuration" and "Autoscaling Group".

Launch Configurations

• Open section Services > EC2 > AUTO SCALING > Launch Configurations and click Create Auto Scaling group then Create launch configuration.
• Choose base image. As we created earlier our preconfigured instance, click "My AMIs" and select image.
• Select instance type and click "Next".
• Give name to launch configuration e.g. "Sample web-app launch configuration".
• Click "Next" then again "Next"
• Select existing security group (created above for web application) and click Review.
• Click Create launch configuration and choose key for ssh access.

Create Auto Scaling Group

• Give name for group
• Select group size (put 2 to have one instance in two different availability zones)
• Choose Subnet (availability zones) let choos a and b zones which were selected for load balancer.
• Go to Advanced Details and select that group will select traffic from one or more load balancer.
• Then select name of target group created for load balancer, and click Next
• Define scaling policy (when and how to scale application), and click Next
• Configure notification as needed and click Next
• Add tegs to ad some metadate to created resource, and click Review.
• Click Create Autoscaling Group for complete process and Close to come back to autoscaling home page.

Removed unused instances

As now we have aouto scaling group we can remove instances created earlier.

TODO: complete aoute scaling configuration step

Reources

• Load balancer Security group http://docs.aws.amazon.com/elasticloadbalancing/latest/classic/elb-security-groups.html#elb-vpc-security-groups
• health check troubleshoot http://docs.aws.amazon.com/elasticloadbalancing/latest/classic/ts-elb-healthcheck.html#ts-elb-healthcheck-failed-vpc
• http://docs.aws.amazon.com/AmazonECS/latest/developerguide/create-application-load-balancer.html#alb-configure-routing
• https://aws.amazon.com/premiumsupport/knowledge-center/troubleshoot-unhealthy-checks-ecs/
• https://docs.oseems.com/general/operatingsystem/linux/automatically-run-program-on-startup
• http://docs.aws.amazon.com/AWSEC2/latest/UserGuide/user-data.html
• http://docs.aws.amazon.com/AmazonECS/latest/developerguide/private-auth.html
• https://docs.docker.com/engine/reference/commandline/login/#login-to-a-self-hosted-registry