Examkin

Question

Difficulty: Very hardSelecting and Designing Compute and Application Platforms

A global logistics organization is architecting an IoT fleet telemetry platform on Google Cloud. The system consists of two core components:

1. An event ingestion web service that receives containerized HTTPS and gRPC telemetry payloads from 500,000 active vehicles. Traffic fluctuates unpredictably throughout the day, requiring rapid scale-to-zero capabilities and minimal operational infrastructure management.
2. A legacy route optimization service written in C++ that requires fine-grained Linux kernel parameter tuning (sysctlsysctl) and shared memory (shmshm) IPC configurations, operating at a continuous baseline load.

Which compute platform architecture meets these technical requirements while minimizing operational management overhead?

  1. Deploy the event ingestion service on Cloud Run, and deploy the C++ route optimization service on Compute Engine Managed Instance Groups (MIGs) utilizing custom virtual machine images.Answer
  2. B
    Deploy both the event ingestion web service and the C++ route optimization service on Google Kubernetes Engine (GKE) Autopilot clusters.
  3. C
    Deploy the event ingestion service on GKE Standard clusters with Cluster Autoscaler, and deploy the C++ route optimization service on Cloud Run as a sidecar container deployment.
  4. D
    Deploy the event ingestion service on App Engine Standard, and deploy the C++ route optimization service on Cloud Functions (2nd gen).

Answer

Deploy the event ingestion service on Cloud Run, and deploy the C++ route optimization service on Compute Engine Managed Instance Groups (MIGs) utilizing custom virtual machine images.
Selecting Cloud Run for the event ingestion service satisfies the containerized HTTP/gRPC scale-to-zero requirement with zero server management overhead. Selecting Compute Engine Managed Instance Groups (MIGs) with custom images for the legacy C++ service allows direct configuration of OS-level parameters, system kernel settings (sysctlsysctl), and shared memory structures that managed container environments restrict.

Step-by-Step Solution

1
Analyze workload 1 (event ingestion) requirements
Identified HTTP/gRPC protocol needs, containerized packaging, rapid auto-scaling from zero, and low operational overhead requirement.
Cloud Run is the optimal Google Cloud compute platform for stateless containerized workloads requiring HTTP/gRPC support and scale-to-zero operational simplicity.
2
Analyze workload 2 (legacy C++ optimization service) requirements
Identified low-level OS dependency requirements including custom sysctlsysctl kernel parameters and shmshm shared memory configuration.
Fully managed compute environments like Cloud Run, App Engine Standard, and GKE Autopilot restrict host kernel modifications. Compute Engine (VMs/MIGs) provides complete OS and kernel control.
3
Synthesize optimal architecture
Combine Cloud Run for serverless ingestion with Compute Engine MIGs for low-level OS kernel customization.
This hybrid platform selection aligns with Google Cloud architectural best practices by choosing serverless abstractions where possible while utilizing IaaS for low-level system dependencies.

Key Concept

Compute Platform Selection Matrix (Cloud Run vs GKE Autopilot vs Compute Engine based on OS/Kernel dependencies and scaling characteristics)
Rate this question

Topics

Selecting and Designing Compute and Application Platforms