Examkin

Question

Difficulty: HardDesigning Infrastructure for Technical Requirements and High Availability

A national healthcare scheduling network is migrating its mission-critical appointments backend to Google Cloud. The application requires global ACID transactions, a recovery point objective (RPO) of 0, a recovery time objective (RTO) under 1 minute, and guaranteed 99.999%99.999\% availability. Additionally, the architecture requires dedicated, SLA-backed hybrid connectivity to on-premises data centers capable of sustaining 20 Gbps of bandwidth without relying on public internet routing. Which TWO architectural decisions should you make to satisfy these technical and high availability requirements?

  1. Deploy a multi-region Cloud Spanner instance configuration to manage transactional backend data across multiple GCP regions.Answer
  2. Provision redundant 10 Gbps Dedicated Interconnect connections across two distinct colocation facilities (edge availability domains).Answer
  3. C
    Configure High Availability Cloud VPN (HA VPN) with 8 encrypted IPsec tunnels deployed across two separate Cloud Routers.
  4. D
    Configure application load balancer health checks to execute a SQL SELECT query directly against the Cloud Spanner database endpoint.
  5. E
    Use VPC Network Peering to transitively route on-premises traffic from a central hub VPC to multiple application spoke VPCs.

Answer

To meet zero RPO, near-zero RTO, 99.999% availability, and high-throughput dedicated hybrid bandwidth, the solution must combine a multi-region Cloud Spanner database deployment with redundant 10 Gbps Dedicated Interconnect connections across two separate edge availability domains.
The combination of multi-region Cloud Spanner and redundant Dedicated Interconnect satisfies all constraints. Multi-region Cloud Spanner provides synchronous global replication with RPO=0, RTO under 1 minute, and a 99.999% SLA. Redundant 10 Gbps Dedicated Interconnect connections deployed across distinct edge locations provide high bandwidth (20 Gbps) and private connectivity backed by Google's enterprise availability SLA.

Step-by-Step Solution

1
Analyze storage high availability and continuity requirements
Cloud Spanner multi-region instance configuration is required because standard Cloud SQL cannot deliver multi-region RPO=0 with active-active global consistency and a 99.999% SLA.
Cloud Spanner uses Paxos consensus and TrueTime to deliver synchronized multi-region replication meeting RPO=0 and sub-minute failover (RTO < 1 min).
2
Evaluate hybrid networking connectivity requirements
Dedicated Interconnect across redundant colocation sites is selected over HA VPN.
HA VPN maxes out at 3 Gbps per tunnel over public lines, whereas 20 Gbps throughput with strict availability SLAs demands Dedicated Interconnect.
3
Eliminate misconfigured architectural patterns
Avoid non-transitive VPC peering reliance and deep dependency health checks.
VPC peering does not extend transitive routes to on-premise interconnects, and database-level load balancer health checks introduce risk of cascading failure.

Key Concept

Designing Enterprise High Availability Architectures for Strict SLA, RPO/RTO, and Hybrid Throughput
Rate this question

Topics

Designing Infrastructure for Technical Requirements and High Availability