Prepare and test your skills

Prepare and test your skills

Project and folder configuration is the setup of Google Cloud projects and folders to create logical environments for billing and administration. A Google Cloud project (opens in a new tab) is a container that holds a set of Google Cloud resources, makes them available for use, and sets the access control and billing rules for those resources. Because those rules attach to the project, the way projects are grouped decides who can manage each workload and how costs are tracked.
Organizing projects by administrative domain means creating a separate Google Cloud project for each country, corporate department, or high-level use case such as edge versus core. This separation keeps management, access control, and billing aligned to the team or purpose that owns the workload. The exact project layout depends on business requirements, so the grouping is chosen to match how environments should be administered.
A Google Cloud project is used for day-to-day management of the deployment inside it:
Hardware ordered within a specific Google Cloud project is always managed in that same Google Cloud project. Cross-project dependencies should be avoided, and resources should be created and used inside each Google Cloud project so each administrative domain stays self-contained.
Applying labels for granular tracking means attaching metadata labels to resources so costs and inventory can be grouped for reporting within and across projects. Resources also have a display_name, and both display_name and labels stay editable after an order is submitted, along with contacts and scheduling fields for specific resource types. Names and labels can therefore be corrected as administrative needs change without recreating the resource. Consistent labels by environment, owner, or use case make tracking and reporting more precise.
Exam tip: Hardware ordered in one Google Cloud project stays managed in that same project — move the management work to the project, not the resource to another project.
Connecting projects to billing means usage reported from services in the project flows to Cloud Billing for cost calculation and invoicing. Cloud Billing computes the cost based on usage, adds that cost to the customer invoice, and invoices are created on month boundaries in US and Canadian Pacific Time. A separate project per administrative domain therefore creates a clear logical environment for billing and administration.
Usage moves from the project to the invoice through an ordered sequence:
The Google Cloud resource hierarchy is a container structure that organizes cloud resources and controls how permissions and policies are inherited. Resource Manager (opens in a new tab) provides this multi-tier hierarchy as organization > folders > projects, where each level is a scope where resources live and where controls can be attached. Administrators group resources logically at a parent level, apply management once, and have those settings flow downward to everything below.
The Organization level container is the top-level container for all projects and resources in Google Cloud. An Organization represents a company or administrative authority and is tied to a Cloud Identity or Google Workspace domain that holds users and groups. Policies, roles, and notification settings defined at the Organization automatically apply to every folder and project beneath it. The Organization is therefore the central point for efficient, company-wide management.
The Folder level container is an optional intermediate layer between the Organization and projects. A Folder groups related projects together, such as by department, team, environment, or application. Because a Folder is a child of the Organization and a parent of projects, any access control or organization policy attached to the Folder is inherited by all projects and resources inside it. Folders are used when one Organization needs separate administrative boundaries without creating multiple Organizations.
The Project level container is the base container where actual cloud resources, such as virtual machines, storage buckets, and databases, are created and used. A Project belongs either directly to an Organization or to a Folder inside an Organization, forming a two-tier path of organization > projects or a multi-tier path of organization > folders > projects. Billing, APIs, access permissions, and policies are scoped to the Project. The Project is therefore the boundary for day-to-day development and cost tracking.
The three levels differ in position and central role:
| Level | Position in hierarchy | Central role |
|---|---|---|
| Organization | Top node above all folders and projects | Defines company-wide administrative authority and central policy |
| Folder | Middle layer between Organization and projects | Groups related projects for delegated management by team or environment |
| Project | Base container under a Folder or Organization | Holds resources and scopes billing, APIs, and workload access |
How centralized management works through inheritance is that attach points and downward flow carry settings from parent to child in Resource Manager. First the administrator attaches Identity and Access Management (IAM) roles, tags, or Organization Policy Service rules at the Organization or Folder, then each child folder, project, and resource inherits them automatically, and only after that can narrower settings be added at lower levels. The Organization Policy Service enforces rules about what can be done, while IAM controls who can do things.
A four-level hierarchy from Organization to Folder to Project to service resources showing IAM allow policies and Organization Policy constraints attached at the top and inherited downward, with effective access as the union of own and ancestor policies.
For example, an Organization policy can restrict the physical locations where resources can be created or prevent the creation of public IP addresses, and Resource Manager defines where that restriction applies. Essential Contacts uses the same cascade, so contacts defined at the Organization automatically receive notifications for all folders and projects below it.
Exam tip: IAM controls who can act on resources, while Organization Policy controls what configuration is allowed on resources.
Policy inheritance is the rule that an access grant or restriction set on a parent container in Google Cloud automatically applies to everything below it. The Google Cloud resource hierarchy (opens in a new tab) flows from the organization node down through folders and projects to individual service resources. A single grant or constraint at the top can therefore propagate to many descendants without repeated configuration.
Container and service resources in the hierarchy are the two kinds of nodes that make up the parent-child tree. Container resources hold other resources and carry policies, while service resources such as Compute Engine instances, disks, subnetworks, and Pub/Sub topics live inside projects and do the actual work. A role granted on a container gives the principal access to both the container itself and the resources inside that container.
The hierarchy has four levels in fixed order:
Traffic for policy evaluation flows from the top down: the organization policy and allow policy apply to folders below it, folder policies apply to projects below them, and project policies apply to service resources inside them.
IAM allow policies and role bindings are the objects that control who can do what on which resource. An Identity and Access Management (IAM) (opens in a new tab) allow policy, formerly called an IAM policy, is a YAML or JSON object attached to a Google Cloud resource. Each allow policy contains a list of role bindings, and each role binding associates one IAM role with one or more principals such as users, Google groups, or service accounts.
Permissions have the form service.resource.verb, such as resourcemanager.projects.list, and they cannot be granted directly to a principal. Principals receive permissions only through roles, which bundle permissions together as predefined roles, customer-created custom roles, or highly permissive basic roles. When an authenticated principal tries an action, IAM checks the resource's allow policy and allows the action only if a binding includes a role with the required permission.
How IAM permissions propagate through inheritance is that setting an allow policy on a container resource also applies that allow policy to all resources in that container. Descendant resources effectively inherit their ancestors' allow policies, so access granted at the organization, folder, or project level flows down to every child project and service resource. For example, if a project's allow policy grants a user the ability to administer Compute Engine VMs, that user can administer any Compute Engine VM in that project regardless of the allow policy set on each VM.
Inheritance creates three practical effects for access design:
roles/iam.securityAdmin) on the organization to manage allow policies everywhere.roles/logging.bucketWriter) on a project to allow writing to a log bucket that has no allow policy of its own.Because inheritance only adds access downward, broad grants at the top have wide impact. Roles should be granted at the smallest scope needed, such as Publisher on a single Pub/Sub topic when only publishing is required, and a shared grant should move to the folder level only when the same user or group needs that role across multiple projects in the folder.
How Organization Policy constraints propagate is that an organization policy administrator enforces a policy at the organization, folder, or project level, and the constraint then applies to that resource and its descendants. The Organization Policy Service provides centralized control through organization policies, which are sets of restrictions called constraints. By default, organization policies are inherited by descendants, so enforcing a policy on a folder enforces it on all projects in that folder.
IAM allow policies and organization policies therefore propagate in the same top-down direction but serve different purposes. An IAM allow policy grants a principal a role that flows from container to contained resources, while an organization policy imposes a constraint that flows from container to contained resources. Both depend on the same hierarchy: the organization sits above folders, folders sit above projects, and projects sit above service resources.
Viewing inherited access and editing rules requires working at the ancestor where the policy was granted, not just on the child being inspected. To see an inherited IAM policy, the viewer needs permission on the ancestor, which can come from predefined roles, custom roles, or roles such as Folder Admin and Project IAM Admin. The three key permissions are:
resourcemanager.organizations.getIamPolicy on the organizationresourcemanager.folders.getIamPolicy on the folderresourcemanager.projects.getIamPolicy on the projectIn the Google Cloud console, a resource's IAM page only shows inherited roles if those roles are grantable on that resource type. Inherited role grants cannot be edited from the child resource where they appear, so the change must be made on the ancestor where the role was originally granted. Audit logs record all setIamPolicy() calls, which makes it possible to trace when an ancestor allow policy was created or changed and how that change propagated to descendants.
Exam tip: An inherited role must be edited where it was granted, not on the child that inherits it, and the console hides inherited roles that are not grantable on the child resource.
Identity and Access Management (IAM) allow policies grant permissions only through role bindings that associate a role with principals, while Organization Policy Service constraints impose restrictions on what configuration is allowed. Both propagate from the organization down through folders and projects to service resources, so a role grant or constraint set on a parent container automatically applies to everything below it.
Policy inheritance means an access grant or restriction set on a parent container automatically applies to everything below it in the order organization to folders to projects to service resources. Because of this downward flow, a role granted on a container gives access to the resources inside that container, a constraint enforced on a folder applies to all projects in that folder, and an inherited role grant must be edited on the ancestor where it was granted rather than on the child.
Separate Google Cloud projects should be created for each country, corporate department, or high-level use case such as edge versus core to keep management, access control, and billing aligned to the team or purpose that owns the workload. Hardware ordered within a specific Google Cloud project is always managed in that same Google Cloud project, so resources should be created and used inside each project to keep each administrative domain self-contained.
An enterprise is establishing its Google Cloud resource hierarchy to manage multiple applications across different environments, such as Development, Staging, and Production. The cloud architecture team needs to establish clear administrative boundaries and centrally apply consistent access control policies across collections of projects.
Which level of the Google Cloud resource hierarchy directly serves as a grouping mechanism for projects to apply shared Identity and Access Management (IAM) policies?
Last updated