info

This repo is created with a sample go application deployed as a microservices in containers using Docker by provisioning an compute engine instance in Google Cloud Platform.

Description

This is a sample Go application which connects to Redis. The app increments a counter on an incoming request.

Setting up the app

• Use make build to compile the binary.
• Set the environment variables:
  • DEMO_APP_ADDR: Address where the app should listen to
  • DEMO_REDIS_ADDR: Address where Redis is running

Tasks

• Create a Dockerfile for the app.

• Create a docker-compose.yml for the app which includes

  • redis service, with the data directory of redis mounted at /data in your VM.
  • app service running with port 8000 exposed to the host.

• Write a bash script that creates and boots Vagrant box with Ubuntu.

• Using Ansible provision the VM to:

  • Setup hostname of VM as demo-ops
  • Create a user demo
  • Configure sysctl for sane defaults. For eg: increasing open files limit. Configure a variety of sysctl settings to make the VM a production grade one.
  • Set the system's timezone to "Asia/Kolkata"
  • Install Docker and Docker-Compose
  • Configure Docker Daemon to have sane defaults. For eg: to keep logs size in check.
  • Deploy the docker-compose.yml in /etc/demo-ops and start the services

---

.

.

Deployment method:

Below is bird's-eye view of this setup:

[Go application & redis] << docker-compose << Docker << Compute Engine << Ansible & Vagrant << Google Cloud Platform.

Dependencies:

• python >= 2.6
• google-auth >= 1.3.0
• ruby
• Vagrant
• Ansible Created and added SSH public keys to GCP project metadata Created service account with a JSON key.

Details of scripts used for deployment:

1. I choose GCP to deploy this application.
2. For server deployment I have two options - any one can be used:
   • One using vagrant - bash script (gcp_instance_vagrant.sh)
   • Other using Ansible - (gcp_instance.yml) - BONUS
3. Provisioning server with ansible - basically configuring server - (gcp_post_deployment.yml)
4. Attached Dockerfile created for application.
5. Attached docker-compose.yml will build image and deploy application as container in the provisioned server (demo-ops).

FINAL OUTPUT:

e_vishnu55@demo-ops:~$ curl http://localhost:8000 
welcome to api 1
e_vishnu55@demo-ops:~$ curl http://localhost:8000 
welcome to api 2
e_vishnu55@demo-ops:~$ curl http://localhost:8000 
welcome to api 3
e_vishnu55@demo-ops:~$