AWS Disaster Recovery part 2

Introduce

This post is a study for the following articles.

• English Disaster Recovery (DR) Architecture on AWS, Part II: Backup and Restore with Rapid Recovery

• Korean AWS 기반 재해 복구(DR) 아키텍처, 2부: 신속한 복구를 위한 백업 및 복원

Prerequisites

This posts is next version of previous post.

• AWS Disaster Recovery part 1
  AWS Architecture Blog | Disaster Recovery (DR) Architecture on AWS, Part I: Strategies for Recovery in the Cloud

Of the four DR strategies in AWS, we cover the Backup & Recovery strategy.

Backup within the AWS Region

All AWS Region is consists of Multi-AZ(Availability Zone).
Multi-AZ strategy supports HA(High Availability).

• Point-in-time Recovery

  • Point-in-time recovery for Amazon DynamoDB

  • Restoring a DB instance to a specified time for RDS

• Versioning Function

  • Amazon Simple Storage Service - Amazon S3

Backup to another AWS Region

In the backup with Multi-AZ Regions, it's recommended that the Origin Region and Backup Region use different AWS Accounts.

Process of DR(Disaster Recovery)

1. Detect when a system failure has occurred and assess the impact.

2. Recover infrastructure and data to get the system back up and running.

3. Failover for delivering HTTP requests to recovered infra.

Step 1 : Detect

RTO includes not only recovery and action time, also includes decision time. Among them, Detect Step can be implemented through automation to reduce RTO.

If you look at the example below, you'll see that it's automated up to the Detect Step.

Step 2 ; Recover

Can recover EBS Volumes, RDS DB instances, etc with the created Backup(step 1). And you can use IaC(Infrastructure as Code) tools such as AWS CloudFormation or CloudDevelopmentKit to accomplish This.

Step 3 : Failover

Failover is the process of bringing the system back to a healthy state by forwarding requests to a newly created infrastructure in another Region or another AZ(Availability Zone).