/revize-ec2-dr

EC2 Disaster Recovery

Overview

  • A Windows Server EC2 instance is always-on Web Server: master instance
  • An existing life cycle management policy is in place creating snapshots from a running ec2 instance
  • The same policy copies the snapshots to another AWS region
  • In another AWS region a slave EC2 instance is created from the master instance snapshot
  • This repository is responsible for triggering a CloudWatch event when the snapshot copy between regions is complete
  • The CloudWatch event will trigger AWS Lambda function
  • Lambda function will launch a new EC2 instance from the latest copied snapshot
  • The Lambda function will then terminate the previously provisioned slave EC2 instance identified by a tag

Out-of-scope

  • Creating a life cycle management policy
  • Configuring EC2 master instance as a Web Server

In-scope

  • Configure CloudWatch event using a bash script and aws-cli
  • Develop a Lambda function triggered by CloudWatch event
  • Deploy AWS lambda function and required artifacts using a bash script and aws-cli
  • Launch a dummy master Windows Server EC2 instance with a C:\hello-world.txt file
  • Manually create a snapshot and copy to another region
  • Demonstrate that a new instance has been created in another region to replace an existing slave instance
  • Login to re-provisioned EC2 slave instance to confirm that C:\hello-world.txt content matches the master instance

Acceptance Criteria

  • Function to use python 3.8 runtime
  • Nothing is hard-coded in the function code, use environment variables instead
  • Function should write its errors to CloudWatch log for troubleshooting purpose