instanceScheduler

What it does ?

Usage

Deploy trough Serverless Application Repository

1. Go to the instance-scheduler repository

2. Click on Deploy

3. Input your email (the email is needed for the resources' status report)

Custom deployment

NOTE: You must have golang or above in order to be able to run this

1. Clone this repository:

   git clone git@github.com:hirvitek/aws-tools.git
   

   and cd to instanceScheduler/

2. Run the script:

   ./deploy.sh <your email>
   

After deployment

• Be sure to check your email to accept the SNS subscription
• Access SSM Parameter store and input your scheduler configuration

Config

You can configure the instance scheduler by changing the SSM Parameter called scheduler-config.

Sample config:

{
	"report": {
		"sendReport": true,
		"hour": 12
	},
	"period": {
		"pattern": "office_hours"
	},
	"timeZone": "Europe/Helsinki",
	"resources": {
		"my-db-instance-123": {
			"type": "RDS"
		},
		"my-ecs-cluster:my-fargate-service": {
			"type": "Fargate"
		}
	}
}

Allowed parameters

Resource types:

• EC2:

  Will stop or start your EC2 instance

• RDS:

  Will bring simply stop or start the RDS specified instance

• AURORA:

  This will simply stop or start your Aurora DB cluster

• Fargate:

  For this type the instance scheduler will only bring the DesiredCount to 0 or 1. The identifier for this resource is: <ecs-cluster-identifier>:<fargate-service-identifier>

Allowed patterns:

• office_hours: resources are awake from 7 am till 5 pm and sleeping during Saturday and Sunday
• permanent_shutdown: always sleeping
• permanent_on: always awake

Road map

• Add ECS and Autoscaling: both will set max size to 0
• Add VPC endpoint: create or destroy VPC endpoints, since their price is a base of 7$ per month, if you have many endpoints is worth to create of destroy them when needed