Examkin

Question

Difficulty: HardDesigning Infrastructure for Technical Requirements and High Availability

A global retail bank is modernizing its core payment authorization service on Google Cloud. The application requires multi-region high availability with an uptime SLA of 99.999%, zero Recovery Point Objective (RPO=0RPO = 0) for transactional ledger data, and low-latency microservice handling. Additionally, the system must establish a sustained 8 Gbps hybrid connection to an on-premises mainframe environment with enterprise SLA guarantees. Which TWO architectural decisions should you recommend to meet these technical requirements? (Select TWO)

  1. Deploy the transactional ledger database on Cloud Spanner configured with a multi-region instance deployment across primary and secondary regions.Answer
  2. Provision a Dedicated Interconnect topology with redundant 10 Gbps VLAN attachments provisioned across separate metro edge locations.Answer
  3. C
    Configure an HA Cloud VPN gateway with four aggregated 3 Gbps Cloud VPN IPsec tunnels per region to handle cross-premises network traffic.
  4. D
    Configure the internal application load balancer health check probes to execute a SQL database query on every check interval to verify downstream dependency health.
  5. E
    Autoscale the microservice compute tier using Compute Engine managed instance groups (MIGs) driven strictly by a target CPU utilization threshold of 80%.

Answer

The architect must deploy Cloud Spanner using a multi-region instance configuration to guarantee RPO=0RPO = 0 and 99.999% availability, and establish a Dedicated Interconnect connection with 10 Gbps VLAN attachments to support sustained 8 Gbps hybrid network traffic with SLA guarantees.
Selecting Cloud Spanner with a multi-region configuration meets the strict 99.999% SLA uptime requirement and guarantees synchronous multi-region replication with RPO=0RPO = 0. Selecting Dedicated Interconnect with 10 Gbps VLAN attachments provides the required physical bandwidth capacity and SLA needed for sustained 8 Gbps cross-premises data transfer.

Step-by-Step Solution

1
Evaluate storage tier high availability and RPO requirements
Cloud Spanner multi-region deployment is selected because it uses Paxos-based synchronous replication to deliver 99.999% availability and zero RPO for transactional data across regional boundaries.
Standard regional Cloud SQL instances offer up to 99.95% SLA and failover latency, failing the multi-region 99.999% SLA and RPO=0RPO = 0 requirement.
2
Evaluate hybrid network bandwidth and SLA requirements
Dedicated Interconnect with redundant 10 Gbps links is selected because it provides dedicated physical infrastructure capability for sustained 8 Gbps traffic.
Cloud VPN is limited to 3 Gbps per tunnel and is unsuited for sustained 8 Gbps enterprise workloads requiring strict SLAs.
3
Review operational stability and autoscaling design choices
Avoid deep dependency health checks and CPU-only autoscaling to prevent cascading failures and scaling bottlenecks under I/O load.
Shallow health checks isolate backend instance health from database state, avoiding widespread un-healthy node removals.

Key Concept

Designing High-Availability Multi-Region Architectures and Dedicated Hybrid Connectivity
Rate this question

Topics

Designing Infrastructure for Technical Requirements and High Availability