Getting started in Amazon Route 53 Application Recovery Controller - Amazon Route 53 Application Recovery Controller

Getting started in Amazon Route 53 Application Recovery Controller

With Amazon Route 53 Application Recovery Controller, you can set up for recovery readiness and create readiness checks for an application, and set up routing controls to reroute traffic for failover. You can also review recommendations to modify your application's architecture for improved resiliency.

To use Route 53 ARC to recover from application failures, you set up at least two application replicas, or cells. Each cell represents an AWS Region or Availability Zone. After you've set up your application resources in cells that align with Availability Zones within a Region, there are some additional steps, listed here, that you must take to make sure that your application follows a recovery-oriented design. After you've implemented these steps, you can use Route 53 ARC for cross-Availability Zone failover.

Tip

To help simplify set up, we provide AWS CloudFormation and HashiCorp Terraform templates that create an application with redundant replicas that fail independently of one another. Learn more and download the templates in Recovery readiness with a new application.

To prepare to use Route 53 ARC, set up a resilient recovery process for your application. The following steps are an overview of how you can prepare your application environment to use Route 53 ARC:

  1. Deploy independent copies of your application stack (networking and compute layer) as standby replicas so that you can fail over traffic across the stacks. You should not have any cross-cell dependencies in the application code where a failure of one cell would impact others. To fail over between Availability Zones or AWS Regions, the boundaries of your cells must align with the isolated infrastructure constructs in AWS (Availability Zones or AWS Regions). However, if you don't set up your boundaries to align with Availability Zones or AWS Regions, Route 53 ARC still supports fail over across cells.

  2. Replicate all required stateful data across the cells. You can use AWS database services to help replicate your data. For example, for high availability, you can add read replicas for Aurora instances across Availability Zones. During failover, you can then promote a replica to be the primary database instance.

  3. Configure each cell to expose a DNS domain name. The domain name must represent the top-level AWS resource in the cell, and act as the front door for servicing client requests to that cell. A top-level resource might be, for example, an Elastic Load Balancing load balancer or an API from API Gateway. Regardless of the resource that you add the name to, the domain name must only direct requests to infrastructure within the cell.

  4. To help you determine the best structure for your application, Route 53 ARC can run an analysis and provide architecture recommendations for you. In the AWS Management Console, provide the application domain name and other information about the application. Route 53 ARC will supply architecture suggestions to help you modify your application to use a well-architected structure that enables fast and complete failover recovery. For more information, see Getting architecture recommendations in Route 53 ARC.

  5. So that traffic failover that uses Route 53 ARC doesn't create data consistency issues, design data reconciliation logic in your failover Region for Regional failovers of stateful applications that require strict consistency.

The following sections include more detailed information about getting started with Route 53 ARC, depending on whether you have an existing application or if you're setting up a new application.

For more information about working with Route 53 ARC, see the following: