Skip to content

README.md

Latest commit

 

History

History
102 lines (76 loc) · 4.32 KB

README.md

File metadata and controls

102 lines (76 loc) · 4.32 KB

README

How to run this project

Requirements

To run this project you will need:

  • dockerhub account
  • configured k8s
  • helm

Architecture

This chart will create a deployment, hpa, service, serviceAccountand secret. alt text

Running the project

For simplicity I've uploaded my image to a public dockerhub rodrigokondo/rails-hello-world:latest, added the rails master key to the chart values and packaged the helm chart in this repo, so you can run helm install [NAME] [CHART], in this case:

helm install hello-world chart-0.1.0.tgz

From scratch

You can do this from scratch as well:

  • (Recommended) Create a new RAILS_MASTER_KEY
  • Build your image docker build . -t [ TAG]
  • Push your image to which image host you prefer
  • If you created a new RAIL_MASTER_KEY update the value in chart/values.yaml, changing rails.masterkey to your master key
  • run helm install helm install [NAME] [CHART]

Terraform

Infrastructure

Using AWS VPC Module. I'm creating a 3 level vpc with public,private and isolated subnets making use of multiple AZs to ensure high availability. And AWS EKS Module to create a EKS cluster on the previously created VPC. alt text

Manage Terraform

Terraform state file

To manage multiple state files. I'm using S3 as a backend and S3 lockfile, and manage which environment I'm working on, by changing backend.tf with a sed to change my the path before initing terraform something like:

terraform {
  backend "s3" {
    bucket = "infrastructure"
    key    = "environments/ENVIRONMENT.tfstate"
    region = "us-east-1"
  }
}

sed -i '' 's/${ENVIRONMENT}/production/g'

terraform {
  backend "s3" {
    bucket = "infrastructure"
    key    = "environments/production.tfstate"
    region = "us-east-1"
  }
}

I've choose this approach to support a pipeline managements of this terraform. Initializing the terraform at each execution of the pipeline and making use of git flow model to know exactly which environment I should change based on the branch I merged the code.

Terraform variables and secrets

Either store variables on environments/${ENVIRONMENT}.tfvars like I'm doing here or store the tfvars on a management tool or S3.

Terraform Testing

To be honest depends on which part I'm testing. if it's something small or that I can recover fast, with minimal impact to the dev team I'm using the staging environment (to me everybody is eligible to break staging dev and ops should be able to test stuff) if it's something totally new/ big change I'm duplicating the infrastructure and testing on a new environment. To me it depends what I'm testing.

Monitoring

One important thing in monitoring to me is centralized monitoring to be able to correlate logs/metrics, so using newrelic, I would leverage all features provided using the APM, infrastructure monitoring, and logs. And to complete the stack define thresholds, opsgenie and slack for alerts and notices.

Key monitoring points,

  • APM and application logs to have a good view of the application.
  • Clustes agents to get metrics and events.
  • Logs from loadbalancers.
  • Notices when new application versions are deployed.
  • Health checkers for all internal tooling.
  • Metrics from databases.
  • Notices when new terraform versions are deployed.

Project

Testing

To test this project on AWS you will need to follow a few steps:

  1. edit terraform/backend.tf adding your backend to it either changing bucket = "my-bucket or removing the s3 backend completely.
  2. Also you can use terraform/init-terraform script to initialize it. From inside terraform/
./init-terraform demo
  1. Run terraform apply -auto-approve -var-file environments/demo.tfvars
  2. Use AWS to grab the kubeconfig from the newly created EKS with aws eks update-kubeconfig --region REGION_CODE --name CLUSTER_NAME in this case: aws eks update-kubeconfig --region us-east-1 --name demo-cluster
  3. Now you can deploy the application. From this project root folder run helm install hello chart/ -f chart/values-eks.yaml
  4. Now you should see a load balancer and accessing it on port :80 you should be able to see the hello world page.