Examkin

Question

Difficulty: HardCapacity Planning and Infrastructure Workload Scaling Optimization

An enterprise genomics research institution runs large-scale batch DNA sequencing jobs on Google Cloud using Compute Engine Managed Instance Groups (MIGs). The workload experiences predictable, massive surges in compute demand during scheduled bi-weekly alignment runs. During past events, auto-scaling instances failed to launch due to sudden regional compute resource limits, and telemetry data transmission experienced severe bottlenecks across hybrid connections. Which TWO architecture and operational strategy changes should the Cloud Architect implement to optimize workload scaling, ensure guaranteed compute capacity, and manage infrastructure constraints during these peak runs? (Select TWO.)

  1. Submit proactive regional vCPU and N2 instance quota increase requests prior to the scheduled alignment runs.Answer
  2. Create zoned Compute Engine Capacity Reservations matching the required instance specification for the duration of the scheduled batch runs.Answer
  3. C
    Migrate the batch sequencing workload to Google Kubernetes Engine (GKE) Autopilot clusters with a Cloud Spanner database backend to bypass compute quota constraints.
  4. D
    Establish an HA VPN tunnel with dynamic BGP routing to stream raw sequence output back to on-premises data centers without dedicated bandwidth planning.

Answer

The correct recommendations are submitting proactive regional vCPU quota increase requests prior to scheduled runs and creating zoned Compute Engine Capacity Reservations to guarantee compute resource availability.
To optimize infrastructure workload scaling and guarantee capacity during predictable surge events, cloud architects must perform proactive capacity planning. Submitting regional quota increase requests well before scheduled events ensures scaling is not blocked by project-level quota caps. Combining quota management with zoned Compute Engine Capacity Reservations ensures the underlying host infrastructure is reserved and immediately available when Managed Instance Groups scale out.

Step-by-Step Solution

1
Analyze capacity failure points
Identified that autoscaling failures stem from regional GCP quota limits and lack of guaranteed compute hardware availability during peak periods.
MIG autoscaling cannot provision instances beyond project quotas or if zonal host capacity is exhausted.
2
Formulate proactive capacity planning strategy
Request quota increases in advance to ensure quota ceiling accommodates peak instance counts, and reserve zoned capacity for guaranteed hardware allocation.
Quota increases resolve API/project limits while capacity reservations guarantee physical compute stock in specified zones.

Key Concept

Capacity Planning, Compute Reservations, and Regional Quota Management
Rate this question

Topics

Capacity Planning and Infrastructure Workload Scaling Optimization