serverless-chaos-monkey

Serverless Chaos Monkey is a serverless implementation of Netflix Chaos Monkey through AWS Lambdas in Python. Present functionality involves two modes of operation:

1. ecs Mode: Allows you to configure chaos monkey to check resiliency of your ecs cluster by randomly terminating any container instance. This is highly useful for micro-services based architectures.
2. autoscaling Mode: This is conventional chaos monkey implementation that randomly terminate ec2 instances in your autoscaling group.

Configuration

Provide your configuration details in config.py file.

1. mode_of_operation: accepted values: 'autoscaling', 'ecs'.

2)autoscaling_group_name: Name of autoscaling group(required if mode_of_operaton= autoscaling)

3)cluster_arn: Arn of ecs cluster (required if mode_of_operation= ecs)

Details of Files

lambda.py : Contains lambda handler function.

config.py : Provide Configuration details here.

chaos.py : Implementation of functions to introduce chaos.

Deployment

Iam Policies : Define a role for lambda and attach following policies:

1. AmazonEC2FullAcess
2. AmazonECS_FullAccess

Create a lambda deployment package and mention lambda handler name as lambda.lambda_handler while deploying the package.