E3S deploy by terraform

Feel free to support the development with a donation for the next improvements.

Prerequisites

AWS Marketplace subscriptions

Zebrunner Selenium Grid Agent

Linux based ECS optimized instance with embedded Zebrunner tuning for scalable and reliable browser images usage

Zebrunner Selenium Grid Agent - Windows

Windows based ECS optimized instance with embedded Zebrunner tuning for scalable and reliable browser images usage

Software

• Installed git
• Installed terraform
• Installed aws cli

AWS resources

• Dynamodb table that will store terraform lock. The table should contain only partition key with name LockID.

Permissions

Configured aws profile with the following policies:

1. In terraform-policy.json should be replaced the next placeholders:

   • {env} - Prefix for almost all e3s aws resources
   • {region} - Aws region in which all e3s resources will be deployed
   • {account} - Aws account id
   • {bucket_name} - Name of a bucket that will store assets (videos, logs, etc.) created by e3s
   • {state_bucket_name} - Bucket to save terraform state
   • {state_bucket_key} - Key of the state file stored in {state_bucket_name}
   • {dynamodb_table} - Dynamodb table name, which will support terraform lock mechanism
   • {dynamodb_region} - Region of the {dynamodb_table}

2. In terraform-ec2-view-policy.json should be replaced the next placeholders:

   • {env} - Prefix for almost all e3s aws resources
   • {region} - Aws region in which all e3s resources will be deployed
   • {account} - Aws account id

3. In terraform-ec2-deploy-policy.json should be replaced the next placeholders:

   • {env} - Prefix for almost all e3s aws resources
   • {region} - Aws region in which all e3s resources will be deployed
   • {account} - Aws account id
   • {e3s-key-name} - Key pair name to attach to e3s-server instance

Deploy steps

1. Clone repository and navigate to deploy folder

git clone https://github.com/zebrunner/e3s-terraform-deploy.git && cd ./e3s-terraform-deploy/deploy

2. Adjust values in terraform.tfvars file

Mandatory vars:

• environment - part of the name of all e3s resources. Most of the resources will have name like e3s-{environment}-<resource-name>.
• region - AWS region in which all e3s resources will be deployed.
• e3s_key_pair_name - AWS key pair name, that will be attached to e3s-server instance.
• bucket - (object) Definition of the bucket where all e3s assets are stored. Contains the following fields:
  • exists - (boolean) determines whether a new bucket should be created or not.
  • name
  • region

Optional vars:

• cert - TLS certificate arn to attach to ELB listener. If none provided, load balancer will only support http protocol.
• e3s_server_instance_type - Instance type of the e3s-server. Default value: m5n.large.
• allow_agent_ssh - Value type: boolean. Default value: false. Allows ssh connection to agent instances by newly created key name only from e3s-server instance.
• enable_cloudwatch - (boolean) Specifies if task logs should be stored in cloudwatch. Default value: false.
• data_layer_remote - (boolean) Determines whether to create rds and elasticache services or use local ones instead. Default value: true.
• nat - (boolean) Determines whether to create private subnets and provide internet connections via nat gateways. Default value: false.
• max_az_number - (number) Determines number of availability zones to use in the specified region. Default value: 3.
• profile - AWS profile to be used by terraform. Default value: default.
• remote_db - (object) Configuration of RDS service credentials. Ignored if data_layer_remote is false. Fields: username and pass. Default value:

  {
      username = "postgres"
      pass     = "postgres"
  }
  
• instance_types = (list of objects) Determines autoscaling group instance types and corresponding weights. Fields: * weight - (number) * instance_type Default value:

  [
      {
          weight        = 1
          instance_type = "c5a.4xlarge"
      },
      {
          weight        = 2
          instance_type = "c5a.8xlarge"
      }
  ]
  
• spot_price - (object) spot price per 1 weight in autoscaling group. If not specified, on-demand instances are used instead. Fields: linux and windows
• zebrunner - (object) Configuration of Zebrunner integration. Fields: host, user, and pass.

3. Create and configure config.{service}.tfbackend file.

Example of s3 backend state configuration exists in deploy folder (config.s3.tfbackend file).

4. Init terraform providers

terraform init -backend-config=config.{service}.tfbackend

For example

terraform init -backend-config=config.s3.tfbackend

5. Deploy

terraform apply

6. [Optional] User policices.

For user's use could be created e3s-manage and e3s-monitor policies.