Setting up a cloud solution environment · Topic 3 of 4

APIs, Quotas, Locations and Asset Inventory

Enabling APIs, observability setup, quotas and increases, product availability by location, networking, Cloud Asset Inventory and Cloud Assist.

4 min

In a nutshell

APIs are enabled per project with gcloud services enable; billing may be required.
Allocation and rate quotas; request increases ahead of time.
Check product, machine type and GPU availability per region and zone.

What it is

A new project is not ready to use until its APIs are enabled, its quotas fit the workload, the chosen locations offer the products needed, basic networking and observability are in place, and the team can see what exists through Cloud Asset Inventory. Gemini Cloud Assist can help analyze those resources.

How it works

Enabling APIs

Most Google Cloud services must be enabled per project before use. Enabling an API can require the project to be linked to a billing account.

gcloud services enable compute.googleapis.com container.googleapis.com
gcloud services list --enabled
gcloud services disable SERVICE_NAME

The role needed is Service Usage Admin (roles/serviceusage.serviceUsageAdmin), or a broader role that includes it.

Quotas

Quotas cap how much of a resource a project can use, protecting both the customer (from runaway costs) and Google Cloud (shared capacity).

Quota type	Example
Allocation quota	Number of vCPUs or GPUs in a region, number of VPC networks
Rate quota	API requests per minute

View quotas on the Quotas page in IAM and Admin (or Cloud Quotas), and request an increase there. Increases are reviewed and are not guaranteed, so request them before a planned event such as a launch.

Exam tip

"Instances fail to start with a quota exceeded error before a big launch" means request a quota increase in advance, not change machine types at the last minute.

Product availability by location

Not every product, machine type or GPU is available in every region or zone. Check before deploying:

gcloud compute regions list
gcloud compute zones list
gcloud compute machine-types list --filter="zone:us-central1-a"
gcloud compute accelerator-types list --filter="zone:europe-west4-a"

The Google Cloud locations documentation lists which products run in which regions.

Setting up cloud networking

New projects usually get a default VPC network in auto mode, with a subnet in each region and permissive default firewall rules. For production, teams normally create a custom mode VPC with planned subnets (covered in the networking pages), and many organizations block default network creation with an organization policy.

Setting up Google Cloud Observability

Cloud Monitoring and Cloud Logging are available in every project. Setup tasks include enabling the APIs, deciding which projects a metrics scope covers (so one project can monitor several), installing the Ops Agent on VMs for OS and application metrics and logs, and creating dashboards and alerting policies.

Cloud Asset Inventory

Cloud Asset Inventory keeps a searchable, time-aware inventory of resources and IAM policies across projects, folders and the organization.

Capability	Use
Search resources and IAM policies	"Which buckets exist in the organization?" "Who has Owner anywhere?"
Export to BigQuery or Cloud Storage	Analyze the inventory with SQL
Asset history	See how a resource's configuration changed over time
Real-time feeds	Send changes to Pub/Sub for alerting

gcloud asset search-all-resources --scope=organizations/ORGANIZATION_ID --asset-types="storage.googleapis.com/Bucket"

Gemini Cloud Assist is an AI assistant in the Google Cloud console that can answer questions about your resources, help analyze configurations and costs, and suggest actions, using natural language.

A worked scenario

A team creates a project for a GPU-based training job in Europe. It enables the Compute Engine API, checks with gcloud compute accelerator-types list which zones offer the needed GPU, sees the regional GPU quota is too low and requests an increase a week before the run, installs the Ops Agent, and later uses Cloud Asset Inventory to confirm no stray GPU VMs remain.

How the exam asks about this

"API not enabled" error. gcloud services enable for that API.
"Plan for more vCPUs before a launch." Request a quota increase in advance.
"Find every resource of a type across the organization." Cloud Asset Inventory.
"Is this GPU available in that zone?" gcloud compute accelerator-types list with a zone filter.

What to remember for the exam

APIs are enabled per project with gcloud services enable; billing may be required.
Allocation and rate quotas; request increases ahead of time.
Check product, machine type and GPU availability per region and zone.
Production uses custom mode VPCs instead of the default network.
Observability setup: APIs, metrics scope, Ops Agent, dashboards, alerts.
Cloud Asset Inventory searches, exports and tracks history of resources and IAM policies.

Watch out

Quota increases are not instant or guaranteed. Waiting until launch day is a classic failure.

Common mistakes

Assuming every region offers every GPU. Check accelerator availability per zone.

Using the default network for production. Plan a custom VPC.

Searching project by project for resources. Cloud Asset Inventory searches the whole organization.

Check yourself

Which command enables the Kubernetes Engine API in a project?
gcloud services enable container.googleapis.com.
Which service searches resources and IAM policies across the whole organization?
Cloud Asset Inventory.
What should you do before a launch that needs many more vCPUs than the current quota?
Request a quota increase in advance.

Topic 3 of 4 in Setting up a cloud solution environment

Test yourself on Setting up a cloud solution environment

The full Google Cloud Associate Cloud Engineer question bank has 661 questions, with timed practice exams and an explanation for every answer.

Subscribe Try Sample Exam

This study guide is independent and is not affiliated with, endorsed by, or sponsored by Google Cloud. Vendor names, certification names and exam codes are trademarks of their respective owners and are used for identification only.