Examkin

Soru

Zorluk: Çok zorHigh Availability and Disaster Recovery Design

A manufacturing corporation is designing a disaster recovery (DR) architecture for its new high-precision IoT telemetry ingestion engine. The ingestion layer runs on Amazon ECS on AWS Fargate in the primary Region (eu-west-1) across three Availability Zones. The telemetry data is stored in an Amazon Aurora PostgreSQL DB cluster in eu-west-1. Ingestion tasks must make outbound API calls to an external validation service using static IP addresses.

The secondary Region for DR is eu-central-1. The corporation demands a recovery point objective (RPO) of less than 1 minute, a recovery time objective (RTO) of less than 15 minutes, and minimal idle infrastructure costs in the secondary Region.

Which two actions should a Solutions Architect take to design a highly available and resilient solution that meets these requirements? (Select TWO.)

  1. Configure an Amazon Aurora Global Database with the primary cluster in eu-west-1 and a secondary cluster in eu-central-1, provisioning a single, low-specification Aurora Replica in the secondary cluster.Cevap
  2. Deploy a public NAT Gateway in each of the three Availability Zones in both the primary and disaster recovery Regions, assign Elastic IP addresses to each NAT Gateway, and whitelist these Elastic IPs with the external validation service.Cevap
  3. C
    Deploy a single public NAT Gateway in one public subnet in both the primary and disaster recovery Regions to route all outbound private traffic, whitelisting the two associated Elastic IP addresses with the external validation service.
  4. D
    Configure hourly snapshots of the Aurora cluster in eu-west-1, copy them to eu-central-1 using AWS Backup, and automate the cluster restoration process during failover.
  5. E
    Configure Aurora Auto Scaling on the secondary cluster in eu-central-1 with a minimum capacity of zero instances, relying on the cluster to automatically scale up write capacity and accept write traffic when client requests are redirected.
  6. F
    Configure Amazon Route 53 Latency-based routing to distribute client telemetry traffic between the Application Load Balancers in both Regions without health checks, allowing Route 53 to redirect client requests during a regional outage.

Cevap

To satisfy the RPO and RTO constraints while avoiding single points of failure, the Solutions Architect should establish an Amazon Aurora Global Database with a single replica in the disaster recovery Region, and deploy NAT Gateways across all three Availability Zones in both the primary and disaster recovery Regions to maintain redundant outbound paths.
Deploying an Amazon Aurora Global Database ensures physical block-level replication with sub-second latency, satisfying the RPO requirement of less than 1 minute. Running a single reader instance in the secondary cluster minimizes ongoing infrastructure costs while facilitating rapid promotion to primary status, which aligns with the RTO of less than 15 minutes. To ensure the ingestion layer remains highly available and resilient to zone failures, public NAT Gateways must be deployed in each Availability Zone in both regions. Whitelisting the Elastic IP addresses of all NAT Gateways guarantees outbound validation API calls succeed from any zone or region.

Adım Adım Çözüm

1
Determine the database replication strategy to satisfy the RPO and RTO constraints.
Select Amazon Aurora Global Database over snapshot replication.
Aurora Global Database replicates data at the physical level with a typical latency of less than a second, meeting the sub-minute RPO. Maintaining a single replica in the secondary cluster provides a low-cost, warm target that can be promoted to a primary cluster in under 15 minutes.
2
Design highly available outbound networking paths for the Fargate tasks.
Provision NAT Gateways in each Availability Zone where the application tasks are deployed.
To satisfy high availability requirements and prevent localized outages from disrupting the ingestion process, each Availability Zone must have its own NAT Gateway. Relying on a single NAT Gateway creates a single point of failure.
3
Configure IP address validation alignment.
Whitelist all Elastic IP addresses assigned to all NAT Gateways across both Regions.
Because tasks run in private subnets and must connect to an external service using static IPs, the Elastic IPs of all NAT Gateways must be pre-whitelisted with the external provider to prevent authentication blocks during a failover.

Anahtar Kavram

Disaster recovery architectures must minimize replication lag to satisfy low RPOs while removing single points of failure in networking paths to prevent local Availability Zone events from cascading into regional service interruptions.
Bu soruyu puanla

Konular

High Availability and Disaster Recovery Design

Bu konudan daha fazla soru