EC2 Crowdstrike Repair (EXPERIMENTAL, NEEDS TESTERS)

This Terraform configuration sets up an Auto Scaling Group (ASG) that runs a PowerShell script to repair unhealthy EC2 instances across all Availability Zones in the current AWS region.

Features

• Launches instances across all available Availability Zones in an ASG, with a single instance per AZ.
• Each instance runs a custom PowerShell repair script, that iterates over all unhealthy instance form the current AZ:

1. detaches the root volume from the current unhealthy instance.
2. attaches it to the current instance.
3. attempts to delete the broken CrowdStrike update file.
4. detaches and re-attaches the volume to the initial instance.
5. restarts the initial instance.

Prerequisites

• Terraform installed (version 0.12 or later)

Usage

1. Clone this repository

2. Initialize the Terraform working directory:

   terraform init

3. Review the planned changes:

   terraform plan

4. Apply the Terraform configuration:

   terraform apply

5. Confirm the changes by typing yes when prompted.

Customization

• To use a different Windows Server version, modify the name parameter in the aws_ssm_parameter data source.
• Adjust the instance_type in the launch template if you need more powerful instances.
• Modify the min_size, max_size, and desired_capacity in the ASG resource if you want a different scaling behavior.

Clean Up

To remove all created resources:

terraform destroy

Contributing

Feel free to submit issues or pull requests if you have suggestions for improvements or bug fixes.

License

This code is (C) 2024 Cristian Magherusan-Stanciu, and released under the MIT license.

Check out more of our OSS repos at LeanerCloud.com.