Examkin

Question

Difficulty: Very hardDeveloping Procedures for Business Continuity and Disaster Recovery Validation

A high-throughput electronic health record (EHR) analytics platform on Google Cloud operates primarily out of us-central1 with an automated failover target in us-east4. The application requires a Recovery Point Objective (RPO) of under 1 minute and a Recovery Time Objective (RTO) of under 15 minutes. During a scheduled disaster recovery (DR) simulation drill, database replication verified clean state synchronization, but application compute instances in us-east4 failed to scale up to accept incoming traffic due to regional vCPU quota exhaustion, leading to an RTO breach. Which procedure should the Cloud Architect mandate to resolve this issue and validate future DR readiness?

  1. Provision Compute Engine capacity reservations in the target DR region for baseline workload requirements and include automated regional quota pre-checks in the DR validation runbook prior to initiating traffic failover.Answer
  2. B
    Rely on standard Compute Engine Managed Instance Group autoscaling in us-east4 and configure Cloud Monitoring alerts to automatically submit urgent GCP quota increase requests at the moment failover is triggered.
  3. C
    Change the disaster recovery architecture from warm standby to a cold standby backup-and-restore model using nightly disk snapshots to eliminate compute quota usage until a disaster occurs.
  4. D
    Replace cross-region Cloud Load Balancing with HA VPN tunnels between the two regions to bypass GCP regional Compute Engine quota constraints during traffic failover.

Answer

Provision Compute Engine capacity reservations in the target DR region for baseline workload requirements and include automated regional quota pre-checks in the DR validation runbook prior to initiating traffic failover.
Reserving capacity in the destination DR region guarantees compute availability when failover occurs. Combining this with automated pre-check validation steps ensures that quotas and capacity are verified before traffic failover is initiated, maintaining strict adherence to RPO and RTO bounds.

Step-by-Step Solution

1
Analyze the root cause of the DR failure
The failure was caused by Compute Engine regional vCPU quota exhaustion during autoscaling in the secondary region (us-east4).
Managed Instance Groups cannot launch instances beyond the project's regional vCPU quota limit regardless of autoscaling rules.
2
Identify the mechanism to guarantee compute availability in target region
Compute Engine Capacity Reservations guarantee instance availability in a specific zone/region.
Reservations ensure that capacity is allocated and ready during a DR event without risking stockouts or quota surprises.
3
Establish validation procedures for DR execution
Incorporate automated API-based quota checking in the DR runbook prior to traffic redirection.
Validating quotas as a pre-flight step in the DR procedure prevents partial failovers and guarantees compliance with the 15-minute RTO target.

Key Concept

Disaster Recovery Capacity Planning and Quota Validation Procedures
Rate this question

Topics

Developing Procedures for Business Continuity and Disaster Recovery Validation