Everyone would like to schedule EC2 instances and save a couple of bucks. Today, I'm bringing you a small Python that can help you. The steps are the following ones:
Step 1. Create a lambda function called: ec2_scheduler
Step 2. Copy and paste the following code:
import boto3 import os import json region = os.environ['REGION'] ec2 = boto3.client('ec2', region_name=region) instances = [] def lambda_handler(event, context): action = event['Action'] response = ec2.describe_instances(Filters=[{'Name' : 'instance-state-name','Values' : [action]}, {'Name': 'tag-key', 'Values': ['auto-scheduled']}]) reservations = response['Reservations'] for reservation in reservations: for instance in reservation['Instances']: instanceId = instance['InstanceId'] for tag in instance['Tags']: if tag['Key'] == 'auto-scheduled' and tag['Value'] == 'true': instances.append(instanceId) if (action == 'stopped'): if (len(instances) > 0): ec2.start_instances(InstanceIds=instances) else: if (len(instances) > 0): ec2.stop_instances(InstanceIds=instances)
Step 3. Set a tag to your EC2 instance called: auto-scheduled that has a value assigned as true.
Step 4. Set the REGION in your environment variables.
Step 5. Add a new Policy in the configuration and permission section of your Lambda
that contains:
{ "Version": "2012-10-17", "Statement": [ { "Effect": "Allow", "Action": [ "logs:CreateLogGroup", "logs:CreateLogStream", "logs:PutLogEvents" ], "Resource": "arn:aws:logs:*:*:*" }, { "Effect": "Allow", "Action": [ "ec2:Start*", "ec2:Stop*", "ec2:Describe*" ], "Resource": "*" } ] }
Step 6. Schedule your Lambda
as triggers with cron expressions like these ones:
- For starting:
cron(0 6 ? * MON-FRI *)
- For stopping:
cron(0 16 ? * MON-FRI *)
Step 7. Set a JSON
that contains the following expression to know if it's starting or stopping:
- For starting:
{ "Action": "stopped" }
- For stopping:
{ "Action": "running" }
And that's all!
Follow me on:
Banner credits:
Top comments (0)