AWS Elastic Disaster Recovery Cheat Sheet

• AWS Elastic Disaster Recovery is a fully managed service providing fast, reliable, and cost-efficient disaster recovery for physical, virtual, and cloud-based Amazon Web Services (AWS) servers.
• It continuously replicates block-level data from the source environment, such as on-premises infrastructure or another cloud provider, to a staging area subnet in AWS. In the event of a disaster or during a drill, AWS DRS enables the rapid launch of fully operational recovery instances in AWS, ensuring minimal downtime and data loss.

Features

• Continuous Replication
  • Block-level replication from on-premises or cloud to AWS.
• Automated Orchestration
  • Automates failover, recovery, and failback operations.
• Non-disruptive Testing
  • Run drills without affecting production environments.

• Point-in-Time Recovery
  • Supports crash-consistent and point-in-time snapshots.
• Broad OS Support
  • Works with Windows, Linux, and various distributions.
• Cross-Region and Cross-AZ Failover
  • Enables regional and zonal disaster recovery within AWS.
• Agent-based Replication
  • A Lightweight agent is installed on source machines.
• Flexible Instance Types
  • You can now launch recovery instances using instance types that differ from your source servers, optimizing for cost or performance during drills and recovery.
• AWS Outposts Support
  • Supports replication and recovery to AWS Outposts, allowing for low-latency disaster recovery on-premises.
• Integration with AWS Services
  • Integrates with CloudWatch, CloudTrail, IAM, and AWS Systems Manager.
  • Agent-based Replication: A lightweight agent is installed on source machines.
  • Supported Sources:
    • On-premises (VMware, Hyper-V, physical servers).
    • AWS EC2 instances.
    • AWS Outposts.
    • Other cloud providers (via custom setup).

Use Cases

• Set up DR for physical or virtual data centers in AWS.
• Protect workloads hosted in other clouds (e.g., Azure, GCP) by replicating to AWS.
• Protect workloads within AWS by replicating EC2 instances to a different region.
• Lift and shift applications into AWS with near-zero downtime.
• Meet RTO/RPO targets for regulatory or business continuity requirements.
• Conduct failover tests without interrupting live workloads.

Best Practices

• Regular Drills:
  • Schedule non-disruptive recovery drills periodically to validate RTO objectives and application functionality.
• Monitor Replication Health:
  • Use Amazon CloudWatch to track replication lag and data transfer bytes. Alert on “Stalled” or “Lagging” states.
• Right-Size Staging Resources:
  • Use the default lightweight staging instances to minimize costs, but ensure your recovery instance types match your performance needs.
• Secure the Staging Area:
  • Apply strict Security Group rules to the staging subnet to prevent unauthorized access to the replication data.
• Validate Quotas:
  • Ensure your target AWS Region has sufficient EC2 and EBS quotas to support the simultaneous launch of all recovery instances during a full disaster.

Elastic Disaster Recovery for Recovery and Failback

• Preparing for failover:
  • Ensure your source servers are healthy and replicating (“Healthy” state).
  • Perform a Recovery Drill to verify that your launch settings (instance types, security groups, subnets) are correctly configured without impacting the source.
• Performing a failover:
  • Initiated when the source site is down or for a scheduled migration.
  • Launches Recovery Instances in the target AWS Region based on a specific Point-in-Time snapshot.
  • You can choose “Initiate recovery job” in the console to spin up these instances.
• Performing a failback:
  • Once the primary site is restored, you can replicate data back from AWS to the original source (on-premises or another region).
  • Requires the Failback Client to reverse the replication direction.
  • Ensures zero data loss during the transition back to the primary site.
• Cross Availability Zone recovery:
  • Protects against AZ failures by replicating EC2 instances to a different Availability Zone within the same Region.
  • Ideal for applications requiring high availability and low-latency recovery without leaving the Region.

Security

• Encryption in Transit and at Rest:
  • Data is encrypted using TLS during transit and optionally with EBS encryption at rest.
• IAM Role-based Access Control:
  • Restrict DRS access using IAM policies and roles.
• VPC Isolation:
  • Replicated instances can be launched into isolated VPCs.
• CloudTrail Logging:
  • All DRS activities are recorded for audit.
• Support for PrivateLink:
  • Use AWS PrivateLink to avoid exposing data to the public internet.

Pricing

• Replication Charge:
  • $0.028 per hour per source server (approx. $20/month).
• Staging Resources:
  • EBS Snapshots: Charged per GB/month.
  • Staging Area EC2: Based on instance type and storage.
• Drill/Recovery Costs:
  • EC2, EBS, and other services used during failover or drills are billed separately.
• No Charge for Failback: Data sent from AWS back to the source is not charged by DRS, but outbound data transfer fees apply.

💡 Tip: You can reduce costs by cleaning up recovery instances and EBS volumes post-drill.

AWS Elastic Disaster Recovery References

What is Elastic Disaster Recovery? – AWS Elastic Disaster Recovery

Disaster Recovery Pricing | AWS Elastic Disaster Recovery | Amazon Web Services

Getting started with AWS Elastic Disaster Recovery – AWS Elastic Disaster Recovery