Davitron / Database_Clustering

A project that demonstrates how to setup a multiple database and loadbalancing them

Database_Clustering

A project that demonstrates how to setup a multiple PostgreSQL databases and loadbalancing them. The project leverages the power of Haproxy as a load balancer. Packer is used to create machine images for Haproxy, NAT, master and 2 slave databases. Google Cloud Provider(GCP) is where the infrastructure runs on.

HOW TO SET UP

• Clone this repo
• Cd into the project
• Run export PROJECT_ID="YOUR_GCP_PROJECT_ID" in the terminal
• Create a folder called secrets in the root directory of this project
• Get your service account key from GCP and place the json file in the folder( Ensure you have already created an account and project before you do so)

create images

• cd into /images/templates
• validate the packer templates with the following commands

   packer validate haproxy.json
   packer validate master.json
   packer validate slave.json
  

• build the images with

    packer build haproxy.json
    packer build master.json
    packer build slave.json
  

Create infrastructure

• Update the variable.tf file with your variables. The 3 images created earlier are also required in the file. The variables consist of the following:
  • images of the Haproxy Load-Balancer, Master and Slave database created with Packer
  • Image type for the NAT instance ( You should use ubuntu-1604-xenial-v20181004 prefarably)
  • IPV4 CIDR for the public and private subnet
  • IP addresses for each instance ( Ensure that they are compatible CIDR specified for the subnet that houses them)
  • Database Password for the master database
  • The machine type (f1-micro seems efficient enough for the project)
  • The region
• Run the following commands

    teraform init
    terraform plan
    terraform apply