Multi-Region Application Architecture

Multi-region application architecture involves deploying application resources (compute, storage, databases) across two or more AWS Regions to achieve higher availability, fault tolerance, and lower latency for global users. On the DVA-C02 exam, you must understand the trade-offs between active-passive and active-active strategies, the use of global services (Route 53, CloudFront, IAM, etc.), and data replication mechanisms.

#### 1. Global Services vs. Regional Services

#### 2. Route 53 – Traffic Routing

Route 53 is the DNS service that routes users to the appropriate regional endpoint. DVA-C02 tests the following routing policies:

Health Checks: Route 53 can monitor the health of endpoints via HTTP/HTTPS/TCP health checks. If a primary region fails health check, failover routing automatically redirects traffic to the secondary region.

Default TTL: DNS record TTL is 300 seconds (5 minutes) by default. Lower TTL (e.g., 60 seconds) allows faster failover but increases DNS query costs.

#### 3. Data Replication Strategies

##### Amazon DynamoDB Global Tables

##### Amazon S3 Cross-Region Replication (CRR)

##### Amazon RDS Cross-Region Read Replicas

##### Amazon Aurora Global Database

#### 4. Compute and Application Layer

##### AWS Lambda – Multi-Region

Lambda functions are regional. To run in multiple regions, deploy the same function code and configuration in each region. Use Route 53 to route API Gateway requests to the appropriate region. For stateful functions, use an external data store (e.g., DynamoDB Global Tables).

##### Amazon ECS/EKS – Multi-Region

Deploy containerized applications using the same task definitions and images across regions. Use a service mesh or Route 53 for traffic routing. For stateful services, use persistent storage like Amazon EFS (which is regional, but can be replicated via EFS replication) or RDS cross-region replicas.

#### Route 53 Failover Routing (AWS CLI)

# Create a health check for the primary endpoint
aws route53 create-health-check --caller-reference "2023-01-01" --health-check-config Type=HTTP, FullyQualifiedDomainName=primary.example.com,Port=80,ResourcePath="/health"

# Create failover DNS records (primary and secondary)
aws route53 change-resource-record-sets --hosted-zone-id ZONEID --change-batch '{
  "Changes": [
    {
      "Action": "CREATE",
      "ResourceRecordSet": {
        "Name": "app.example.com",
        "Type": "A",
        "SetIdentifier": "primary",
        "Failover": "PRIMARY",
        "HealthCheckId": "CHECKID",
        "TTL": 60,
        "ResourceRecords": [{"Value": "1.2.3.4"}]
      }
    },
    {
      "Action": "CREATE",
      "ResourceRecordSet": {
        "Name": "app.example.com",
        "Type": "A",
        "SetIdentifier": "secondary",
        "Failover": "SECONDARY",
        "TTL": 60,
        "ResourceRecords": [{"Value": "5.6.7.8"}]
      }
    }
  ]
}'

#### Enable S3 Cross-Region Replication

# Enable versioning on source and destination buckets
aws s3api put-bucket-versioning --bucket source-bucket --versioning-configuration Status=Enabled
aws s3api put-bucket-versioning --bucket dest-bucket --versioning-configuration Status=Enabled

# Create replication role
aws iam create-role --role-name replication-role --assume-role-policy-document file://trust-policy.json

# Create replication configuration
aws s3api put-bucket-replication --bucket source-bucket --replication-configuration file://replication-config.json

Example replication-config.json:

{
  "Role": "arn:aws:iam::ACCOUNTID:role/replication-role",
  "Rules": [
    {
      "Status": "Enabled",
      "Priority": 1,
      "DeleteMarkerReplication": { "Status": "Disabled" },
      "Filter": { "Prefix": "" },
      "Destination": {
        "Bucket": "arn:aws:s3:::dest-bucket",
        "StorageClass": "STANDARD"
      }
    }
  ]
}

#### Verify DynamoDB Global Table status

aws dynamodb describe-global-table --global-table-name my-table

Scenario 1: Global SaaS Application A SaaS company provides a project management tool to customers worldwide. They deploy an active-active architecture across us-east-1 (N. Virginia) and eu-west-1 (Ireland). Compute runs on ECS Fargate behind Application Load Balancers. User session data is stored in DynamoDB Global Tables, so writes in either region are replicated within seconds. File attachments are stored in S3 with CRR. Route 53 latency-based routing directs users to the nearest region. During a major storm in the US East Coast, us-east-1 becomes degraded. Route 53 automatically routes all traffic to eu-west-1. Users experience slightly higher latency but no data loss because DynamoDB and S3 are replicated. The company meets its RTO of 5 minutes and RPO of 1 second.

Scenario 2: Financial Services Disaster Recovery A bank requires strict regulatory compliance: data must remain in the US, but they need DR. They choose an active-passive architecture with primary in us-east-1 and standby in us-west-2. They use RDS for MySQL with cross-region read replicas. The standby replica is promoted manually during a disaster. S3 CRR replicates transaction logs. Route 53 failover routing with health checks monitors the primary ALB endpoint. The bank conducts quarterly failover drills, promoting the read replica to a standalone instance and updating the application configuration. They observe an RTO of 30 minutes (manual promotion time) and RPO of 5 minutes (replication lag). Misconfiguration of security groups in the standby region once caused connectivity issues during a drill, highlighting the need for consistent infrastructure.

Scenario 3: Media Streaming Platform A video streaming service uses CloudFront as a CDN with multiple origins in different regions. The backend is a serverless application using API Gateway and Lambda, deployed in us-east-1, eu-west-1, and ap-southeast-1. User profiles are stored in DynamoDB Global Tables. Video metadata is stored in S3 with CRR. They use Route 53 geolocation routing to ensure users are directed to the region that contains their licensed content. During a regional outage, CloudFront automatically fails over to a healthy origin. The main challenge is managing replication of large video files; they use S3 Transfer Acceleration to speed up CRR. They monitor replication metrics and have alerts for lag.