Examkin

Question

Difficulty: MediumDisaster Recovery and Business Continuity Execution

A enterprise SaaS payroll platform runs its primary application tier on Compute Engine Managed Instance Groups (MIGs) and its transactional database on Cloud SQL for PostgreSQL in the primary region `us-east1`. The business continuity requirement specifies a Recovery Point Objective (RPO) of 15 minutes and a Recovery Time Objective (RTO) of 2 hours, with strict budget limits preventing a full multi-region active-active deployment. During a scheduled disaster recovery failover exercise to the secondary region `us-central1`, the database read replica was successfully promoted, but compute instance creation in `us-central1` failed due to insufficient regional vCPU quota limits. Which architectural solution should the cloud architect implement to guarantee successful disaster recovery execution while adhering to the RPO, RTO, and cost constraints?

  1. Pre-request and maintain sufficient regional vCPU quotas in the secondary region, and deploy a warm-standby MIG with a minimal baseline instance count that scales out when Cloud DNS failover triggers.Answer
  2. B
    Migrate the underlying relational database to multi-region Cloud Spanner to handle database and application failovers automatically without requiring regional quota planning.
  3. C
    Switch to a cold backup strategy using daily Cloud Storage export snapshots, recreating compute and database resources on demand in `us-central1` via automated Terraform pipelines during failover.
  4. D
    Keep all compute instances turned off in `us-central1` until failover occurs, relying on GCP automated real-time quota adjustment scripts executed during a disaster event.

Answer

Pre-request and maintain sufficient regional vCPU quotas in the secondary region, and deploy a warm-standby MIG with a minimal baseline instance count that scales out when Cloud DNS failover triggers.
The correct answer addresses both the operational bottleneck (vCPU quota limits) and the business constraints (15-min RPO, 2-hr RTO, cost limits). Pre-requesting regional quotas guarantees that GCP will permit scaling in the target region during an emergency. Maintaining a minimal warm-standby Managed Instance Group keeps compute costs low during normal operation while allowing rapid auto-scaling upon Cloud DNS failover.

Step-by-Step Solution

1
Analyze DR Constraints
Identified RPO = 15 minutes, RTO = 2 hours, and strict cost boundaries preventing active-active deployment.
Determines the appropriate disaster recovery pattern (Warm Standby / Pilot Light).
2
Identify Cause of Failover Failure
Compute Engine instance creation failed in the target region due to insufficient vCPU quotas.
GCP quotas are per-region; expanding workloads rapidly in a target DR region during failover requires pre-allocated quota buffer.
3
Select Correct Remediation Architecture
Maintain a warm-standby MIG with minimal capacity and pre-approved vCPU quotas in the secondary region.
Satisfies the 2-hour RTO by permitting rapid scaling while preventing unexpected deployment failures due to quota exhaustion.

Key Concept

Disaster Recovery Execution and Regional Quota Management
Rate this question

Topics

Disaster Recovery and Business Continuity Execution

More questions from this topic