Examkin

Question

Difficulty: HardCapacity Planning and Infrastructure Workload Scaling Optimization

A global online gaming studio hosts a competitive multiplayer platform on Google Cloud using Compute Engine Managed Instance Groups (MIGs) positioned behind regional External Application Load Balancers. The workload experiences predictable, extreme traffic surges during scheduled weekend global tournaments. During previous events, standard reactive autoscaling failed to provision compute instances quickly enough to absorb the rapid spike in player connections, and provisioning operations failed due to regional compute resource limitations. As the Lead Cloud Architect, you need to implement a proactive capacity planning and workload scaling strategy to ensure seamless scalability and prevent resource exhaustion during upcoming tournaments. Which TWO actions should you take to achieve this objective?

  1. Configure schedule-based autoscaling policy rules on the Managed Instance Groups to scale out baseline instance capacity in advance of known tournament start times.Answer
  2. B
    Rely on Compute Engine dynamic autoscaling to automatically adjust regional project vCPU quotas in real time as incoming traffic scales up.
  3. Perform proactive capacity planning by evaluating historical usage peak metrics and requesting regional vCPU quota increases well in advance of scheduled events.Answer
  4. D
    Configure the Managed Instance Groups to autoscale exclusively based on CPU utilization targeted at a high 95% threshold to maximize resource efficiency.

Answer

To optimize infrastructure capacity and prevent service disruption during predictable high-volume events, cloud architects should establish schedule-based autoscaling rules to pre-provision compute instances prior to peak windows and conduct proactive quota management by requesting regional vCPU quota increases in advance.
Combining schedule-based autoscaling with proactive quota management addresses both compute initialization delays and infrastructure quota limits. Schedule-based scaling pre-provisions instance capacity before predictable surges occur, ensuring zero cold-start latency for incoming users. Proactively auditing and submitting regional quota increase requests ensures that GCP project limits do not block the Managed Instance Group from expanding to its required peak instance count.

Step-by-Step Solution

1
Analyze workload surge patterns and response latency of reactive scaling
Identified that reactive metric thresholds (such as CPU or request rate) lag behind instantaneous connection spikes, leading to cold-start delays during event launches.
Predictable surges require pre-allocation of compute capacity to ensure instances are fully initialized and healthy before traffic arrives.
2
Implement schedule-based autoscaling rules on Compute Engine MIGs
MIGs automatically scale up to required minimum baseline instance counts prior to the tournament start time.
Schedule-based scaling rules complement metric-based policies by absorbing initial burst traffic smoothly without initialization bottlenecks.
3
Evaluate regional resource quotas against projected peak load
Determined required regional vCPU, memory, and IP allocation capacity needed for peak instance counts during the tournament.
Autoscaling policies cannot exceed project-level regional quotas. Requesting quota increases prior to deployment prevents provisioning failures.

Key Concept

Proactive Workload Capacity Planning and Scheduled Autoscaling
Rate this question

Topics

Capacity Planning and Infrastructure Workload Scaling Optimization