DR & Migrations Overview

Content	page
AWS DMS	AWS DMS
AWS Backup	AWS Backup
AWS Application Discovery Service	AWS Application Discovery Service
AWS Application Migration Service	AWS Application Migration Service
VMC on AWS	VMC on AWS

🧠 What Is Disaster Recovery (DR)?

Disaster Recovery (DR) is the practice of planning for and recovering from unexpected events that disrupt business operations — like hardware failures, cyberattacks, accidental deletions, natural disasters, etc.

In the cloud, DR focuses on resilience, automation, and cost-effective replication of critical systems and data.

---

📏 RTO and RPO: The Core Metrics

Metric	Definition
RTO (Recovery Time Objective)	How quickly you must restore service after a disruption
RPO (Recovery Point Objective)	How much data loss (time window) you can tolerate

For example:

• If RTO = 4 hours → system must be back online within 4 hours

• If RPO = 15 minutes → you can lose at most 15 minutes of data

---

🧪 Common RTO/RPO Requirements by Industry

Industry	RTO	RPO
Banking	≤ 1 hour	≤ 5 minutes
Healthcare	1–2 hours	≤ 15 mins
SaaS Startups	4–8 hours	1–4 hours
Non-critical apps	24–48 hours	12–24 hours

---

🏗️ Disaster Recovery Strategies in AWS

AWS offers 4 standard DR architectures:

Strategy	RTO	RPO	Cost	Description
Backup & Restore	Hours	Hours	Low	Backup data & config, restore manually after disaster
Pilot Light	Minutes	<1 hour	Medium	Core services (DB, AMIs) always running, others launched on fail
Warm Standby	<30 min	<30 min	High	Scaled-down version always running, scaled up when needed
Multi-site (Hot)	Seconds	Seconds	Very High	Fully duplicated system in multiple regions or AZs

---

🔁 Migration vs DR

Aspect	Migration	Disaster Recovery
Purpose	Move workload permanently	Restore workload temporarily
Involves Cutover	✅ Yes	❌ Not unless disaster occurs
Downtime allowed	Often scheduled	Must be minimized
Data replication	One-time or phased	Continuous or periodic
Tooling	AWS DMS, SMS, Application Migration Service	Snapshots, Replication, CloudEndure

---

☁️ AWS Services for DR

Category	Services	Description
Compute	EC2 AMIs, Auto Scaling	Pre-baked backups, scale after restore
Storage	EBS snapshots, S3 versioning	Durable backups
Database	RDS Multi-AZ, Aurora Global	Automatic failover, cross-region replicas
DNS	Route 53	Health checks, failover routing
Replication	AWS DMS, CloudEndure	Live or scheduled data replication
Automation	Lambda, CloudFormation	Automate failover & restore

---

🔄 Sample DR Workflow: Backup & Restore

1. Use AWS Backup to schedule daily EBS and RDS backups

2. Replicate backups across regions

3. Store application configs (env vars, IAM, templates) in S3 or SSM

4. When disaster occurs:

   • Spin up EC2 instances using AMIs

   • Restore RDS from snapshot

   • Repoint DNS via Route 53

---

💡 DR Best Practices & Tips

Category	Tip
Automation	Use CloudFormation + Lambda to automate restoration & infra setup
Immutable Infra	Use AMIs and containers for fast deployments
Testing	Perform regular DR drills using sandbox accounts or test AZs
Backups	Enable versioning for S3, schedule backups for EBS, RDS
Encryption	Use KMS to secure backups and replicas
Monitoring	Use CloudWatch + SNS for outage alerts and triggering failover
Documentation	Maintain clear SOPs (Standard Operating Procedures) for DR

---

📦 Tooling for Migrations & DR

Tool/Service	Use Case
AWS DMS	Migrate live databases
AWS CloudEndure / MGN	Lift-and-shift of full apps/VMs
AWS Backup	Scheduled backup + cross-region copy
S3 Replication	DR for object storage
RDS Multi-AZ / Read Replica	Hot standby for databases
Route 53 Failover	DNS-based failover
Step Functions	Recovery workflows orchestration

---

🧱 Terraform Tip for Cross-Region DR (S3 Example)

resource "aws_s3_bucket" "primary" {
  bucket = "my-primary-bucket"
  versioning {
    enabled = true
  }
  replication_configuration {
    role = aws_iam_role.replication.arn
    rules {
      id     = "replication-rule"
      status = "Enabled"
      destination {
        bucket        = aws_s3_bucket.secondary.arn
        storage_class = "STANDARD"
      }
    }
  }
}

---

✅ TL;DR Summary

Term	Meaning
RTO	Max acceptable downtime (e.g., 1 hr)
RPO	Max acceptable data loss (e.g., 15 min of data)
DR Plan Types	Backup-Restore, Pilot Light, Warm Standby, Multi-site
Tools	DMS, MGN, S3 replication, Route 53, CloudFormation
Key Advice	Automate everything, test regularly, document recovery steps

---