GCP Professional Cloud Architect Practice Question

An enterprise architecture team must deploy Apigee X for a global retail platform. Requirements: (1) production APIs need low-latency ingress in us-central1 and europe-west1; (2) non-production traffic must be fully isolated from production; (3) each environment must scale automatically during peak sales events; (4) the design should minimize platform cost and stay within the ten-instance limit that Google Cloud enforces per organization. Which deployment plan meets all requirements?

Create three Apigee instances in each required region (dev, staging, prod) for a total of six, and set both the minimum and maximum node count to 6 on every instance to guarantee capacity.
Create two Apigee instances, one in us-central1 and one in europe-west1, and attach the production environment to both through an environment group. Create a third instance in us-central1 that hosts shared staging and development environments. Configure each instance with a minimum of 2 and a maximum of 6 runtime nodes so Apigee X can autoscale.
Provision 12 Apigee instances-one per microservice per region-and use manual scripts to add or remove runtime nodes as load changes.
Use a single multi-region Apigee instance with separate environments for dev, staging, and prod, relying on environment-level quotas to protect production traffic.

Report Issue

Answer Description

Creating two production Apigee X instances-one in us-central1 and one in europe-west1-meets the low-latency ingress goal. A third, separate instance for all non-production environments keeps test traffic off the production runtimes yet remains well under the ten-instance quota. Setting a minimum of two nodes and a maximum of six nodes on every instance lets Apigee X add or remove runtime nodes automatically in response to traffic, containing cost during quiet periods.

The other options fail at least one requirement:

Six instances with the minimum and maximum node count both set to six cannot scale beyond that level and incur higher baseline cost.
A single multi-region instance would share runtime capacity between production and non-production traffic, so isolation is not achieved even though the instance itself is regionally redundant.
Provisioning twelve instances exceeds the organizational limit.

Ask Bash

Bash is our AI bot, trained to help you pass your exam. AI Generated Content may display inaccurate information, always double-check anything important.