AWS re:Invent 2025 - Disaster Recovery (DR) with AWS Elastic Disaster Recovery Service (COP356)

AWS re:Invent 2025 - Disaster Recovery (DR) with AWS Elastic Disaster Recovery Service (COP356)

The Importance of Disaster Recovery (DR)

• Real-life examples of the high cost of downtime:
  • A global company lost $8.2 million in revenue due to an outage
  • A leading US airline lost $150 million in profit due to an outage
  • IDC reports that Fortune 1000 companies can lose $500K to $1 million per year due to application downtime, and 76% had an outage in the past 2 years
• Implementing a DR solution can provide a 12-20% increase in profit margins by protecting the bottom line and gaining a competitive advantage

Choosing the Right DR Strategy

• Two critical metrics to consider:
  • Recovery Point Objective (RPO): How much data can you afford to lose?
  • Recovery Time Objective (RTO): How quickly do you need your service/application back online?
• Three main DR strategies:
  1. Backup: RPO and RTO in hours, most cost-effective
  2. Disaster Recovery: RPO in seconds, RTO in minutes, moderate cost
  3. High Availability: Near real-time RPO and RTO, significant investment for critical applications

AWS Elastic Disaster Recovery Service (AWS DRS)

• Agent-based replication service that continuously replicates data to a lightweight staging area in AWS
• Supports various source environments: EC2 instances, on-premises servers, other cloud platforms
• Allows for easy testing of DR capabilities and fast failover to AWS in the event of a disaster
• Key benefits:
  • Low cost due to efficient replication and no third-party fees
  • Highly automated conversion and recovery process
  • Point-in-time recovery options for ransomware protection
  • Easy testing with one-click drills

Technical Architecture and Pricing

• On-premises to AWS architecture:
  • Replication agent installed on source servers, continuously replicates data to AWS staging area
  • Recovery instances launched in AWS, mapped to source servers
  • Failback process to return to primary site when ready
• Region-to-region DR within AWS:
  • Leverages native AWS services like S3 cross-region replication for efficient data transfer
  • EC2-to-EC2 transfer for low latency
• Pricing:
  • $20/month per server for the service
  • Charges for staging area usage (EC2 and EBS)
  • Temporary failover instance costs during disaster recovery

Advanced Features

• Post-launch actions: Automatically execute predefined scripts and validations on recovered instances
• Systems Manager integration: Automate scripts and commands on recovered instances for connectivity checks, permission validation, etc.

Key Takeaways

• Disaster recovery is critical to protect against the high costs of downtime, with real-world examples of millions in losses
• AWS DRS provides a cost-effective, highly automated, and resilient DR solution that can be tailored to different RPO and RTO requirements
• Advanced features like post-launch actions and Systems Manager integration further streamline the DR process and reduce complexity
• Implementing a robust DR strategy can provide a significant competitive advantage by protecting the bottom line and ensuring business continuity