Question

Difficulty: MediumPlanning Google Kubernetes Engine (GKE) Cluster Architectures

An e-commerce company is architecting a containerized application platform on Google Kubernetes Engine (GKE). The solution consists of two distinct workloads: a stateless, fault-tolerant log analytics pipeline and a legacy core service that requires custom Linux kernel sysctl parameters and specialized node-level privilege access to the host OS. Which TWO architectural decisions should the engineering team implement to satisfy these operational and technical requirements? (Select TWO options.)

  1. Provision a dedicated GKE Standard node pool configured with Spot VMs to execute the fault-tolerant log analytics workload at reduced cost.Answer
  2. Deploy the legacy core service on a GKE Standard cluster to support custom kernel sysctl parameters and host node operating system access.Answer
  3. C
    Deploy the legacy core service on a GKE Autopilot cluster to eliminate node management overhead while applying custom kernel sysctl modifications.
  4. D
    Configure the Horizontal Pod Autoscaler (HPA) to automatically provision additional Compute Engine virtual machine instances during traffic spikes.
  5. E
    Use Spot VM node pools as the primary hosting environment for mission-critical transactional databases to optimize infrastructure compute expenses.

Answer

The correct architecture requires provisioning a GKE Standard node pool using Spot VMs for stateless, fault-tolerant workloads to save costs, and deploying workloads requiring custom Linux kernel sysctl parameters or direct host OS access onto GKE Standard mode.
Deploying on GKE Standard is necessary whenever custom Linux kernel parameters or privileged host OS access is required. Additionally, using Spot VM node pools for stateless log analytics provides significant cost reduction while satisfying the fault-tolerant nature of the processing pipeline.

Step-by-Step Solution

1
Analyze workload requirements for node OS customizations and host access.
GKE Autopilot locks down node configuration for security and operational ease. Therefore, custom sysctl settings and host OS access necessitate GKE Standard mode.
GKE Standard retains cluster node admin control required for OS-level tuning.
2
Analyze cost optimization for the fault-tolerant, stateless log analytics pipeline.
Spot VMs provide significant cost discounts for workloads that can handle sudden node termination.
Stateless log analytics workloads can tolerate preemptions without data corruption.

Key Concept

Selecting between GKE Autopilot vs. Standard modes and identifying appropriate use cases for Spot VM node pools.
Estimated Time:2m 0s
Rate this question