Prepare and test your skills
Prepare and test your skills
Free sample question, answered and explained below. The Professional Cloud Architect practice bank has 2,335 more like it. Try 10 free: instant score, no account.
An organization is designing an automated, event-driven machine learning workflow on Google Cloud. The architecture must satisfy the following technical requirements:
Which architectural approach should you implement?
This approach uses Cloud Scheduler on a fixed time interval (cron) to launch a Vertex AI Batch Prediction job using NVIDIA Triton Inference Server containers.
FLEX_START with configurable max_wait_duration on training worker pools.The objective was to perform scheduled, offline inference over tabular or unstructured datasets using an already trained model.
gsutil and executes local Python training scripts.This approach deploys a dedicated Compute Engine virtual machine configured with physical GPU accelerators that runs a continuous background script (daemon) to check Cloud Storage for new datasets.
FLEX_START capabilities or serverless preemptible quota models.The workload required a persistent, interactive development environment where developers constantly fine-tune models interactively on a single static server.
FLEX_START scheduling strategy and max_wait_duration, then deploys the resulting model artifact to a Vertex AI Endpoint upon completion.This architecture combines Eventarc, Cloud Run functions, and Vertex AI Custom Training using the Dynamic Workload Scheduler (DWS) with the FLEX_START scheduling strategy to create an automated, resilient, serverless machine learning orchestration pipeline.
scheduling.strategy to FLEX_START and defining a max_wait_duration, Vertex AI queues the job until the requested accelerator resources (such as NVIDIA A100 or H100 GPUs) become available, consuming preemptible quota allocation while running the job to completion reliably.
This configuration proposes using Google Kubernetes Engine (GKE) Autopilot paired with Cloud Logging log sinks to handle incoming data events and attempt model registration directly.
FLEX_START scheduler for queuing and managing scarce GPU training capacity.The workload was centered around Kubernetes-native microservices requiring real-time log ingestion and analysis rather than an automated Vertex AI training workflow.
Answer them yourself and get your score the moment you finish. No account needed.
Full access includes
Keep the momentum going with these hand-picked practice scenarios
Last updated