Prepare and test your skills

Prepare and test your skills

Creating a resource hierarchy is the setup of Google Cloud projects, folders, and an organization into a parent-child tree for billing and administration. The resource hierarchy (opens in a new tab) flows in fixed order from organization to folders to projects to service resources, and settings attached at a parent flow downward to everything below it. A separate Google Cloud project is typically created for each administrative domain, such as a department, country, or use case, so management and costs stay aligned to the owner.
A top-down hierarchy from Organization to Folders to Projects to service resources, showing IAM and Organization Policy set at a parent flowing down to all children.
The Organization container is the top-level root tied to a Cloud Identity or Google Workspace domain, and policies set there apply to every folder and project beneath it. A Folder is an optional middle layer that groups related projects by team, environment, or application for delegated management. A Project is the base container where virtual machines, buckets, and databases actually live, and billing, APIs, and access permissions are scoped to that project:
| Level | Position | Central role |
|---|---|---|
| Organization | Top node above all folders and projects | Company-wide authority and central policy |
| Folder | Between organization and projects | Groups projects for team or environment management |
| Project | Base container under folder or organization | Holds resources and scopes billing and access |
How centralized management works is that an administrator attaches Identity and Access Management (IAM) roles, tags, or Organization Policy rules at the Organization or Folder, then each child inherits them automatically. IAM controls who can act, while Organization Policy controls what configuration is allowed, such as blocking public IP addresses or restricting resource locations. Labels and display_name provide granular tracking within and across projects and stay editable after creation, so costs can be grouped by environment or owner without recreating resources.
Exam tip: IAM controls who can act on resources, while Organization Policy controls what configuration is allowed on resources.
Continue with this study guide: Creating a resource hierarchy
Applying organizational policies to the resource hierarchy means enforcing centralized rules that flow from a parent to all of its descendants. The Organization Policy Service (opens in a new tab) provides this control through organization policies made of constraints, which are individual restrictions on operations or field values. When a user tries to create or update a resource that violates an enforced constraint, Google Cloud blocks the operation.
Boolean constraints are on-or-off rules with an enforced state of true or false and no list of values to manage. List constraints permit a request only when its value matches allowedValues and is not denied. Three common list constraints are gcp.resourceLocations for allowed regions, gcp.restrictServiceUsage for allowed API services, and gcp.restrictNonCmekServices for services that require customer-managed encryption keys (CMEK).
Policy inheritance means a policy enforced on an organization covers all folders and projects under it, and a policy enforced on a folder covers all projects in that folder, including projects added later. A policy override is a different policy set directly on a child folder or project, so only that branch operates under the exception while production keeps the strict baseline. First the strict rule is enforced at the top, then the exception is set on the development or testing folder or project where it is justified.
Exam tip: A policy set on a folder applies to every current and future project in that folder; an exception requires a different policy set directly on the lower-level folder or project.
Continue with this study guide: Applying organizational policies to the resource hierarchy
Granting members IAM roles within a project is the process of binding a principal to a role on a specific resource to control what that identity can do. Identity and Access Management (IAM) (opens in a new tab) enforces access through an allow policy that connects a principal, a role, and a resource. A role granted on a project, folder, or organization inherits downward to all contained resources.
A principal is any identity that can receive access, including Google Accounts for individuals, Google Groups for teams, service accounts for workloads, and workforce identity pool identities for external users. In gcloud, a binding uses the form user:my-user@example.com, while the console accepts the plain email address in the New principals field. Roles differ in breadth and maintenance:
| Role type | Who defines it | Scope of access |
|---|---|---|
| Primitive, such as Owner and Editor | Very broad access across all services in a project | |
Predefined, such as roles/SERVICE.IDENTIFIER | Narrow access for a specific service or job | |
Custom, such as projects/PROJECT_ID/roles/IDENTIFIER | Customer organization | Exact permission set for a specific task |
Primitive roles Owner, Editor, and Viewer grant far more access than most tasks need and should not be used in production. Predefined roles support least privilege better because they match a service or job function, and custom roles match a task most closely but must be maintained over time.
The Google Cloud console and gcloud CLI update the same allow policy, so either tool grants the same access. In the console, the workflow is to open IAM for the selected project, folder, or organization, enter the principal identifier exactly, select a least-privilege role with an optional condition, and click Save. In the CLI, the pattern names the resource type, resource, member, and role, such as gcloud projects add-iam-policy-binding my-project --member=user:my-user@example.com --role=roles/cloudbuild.builds.viewer.
Exam tip: In production, do not grant Owner, Editor, or Viewer; grant a predefined or custom role that meets the need.
Managing users and groups in Cloud Identity means creating organization-controlled identities that IAM can then use for Google Cloud access. Cloud Identity (opens in a new tab) is an Identity as a Service (IDaaS) solution that centrally manages users and groups in a domain and can federate with providers such as Active Directory. Each Google Workspace or Cloud Identity account links to exactly one organization resource, and IAM roles granted at that organization flow down to all child projects.
A Google group is a collection of users where each member inherits the IAM roles granted to the group, so access is managed by changing membership instead of editing many bindings. To create a group, open the Groups page for the organization, click Create, enter name and email, add members, and submit. Group creation privileges come from Workspace admin roles, while viewing groups from the Google Cloud console also needs IAM roles such as Organization Viewer. Before deleting a group, revoke all IAM roles from it and wait at least 7 days, because deletion is irreversible.
Google Cloud Directory Sync (GCDS) copies users and groups from Active Directory or LDAP into Cloud Identity using LDAP on the on-premises side and HTTPS with the Directory API on the cloud side. Data flows from Active Directory to GCDS through LDAP, then from GCDS to Cloud Identity, keeping cloud accounts aligned without manual entry. Federation with sync requires two accounts per user, an external account plus a Cloud Identity account, while Workforce Identity Federation needs only the external account.
Exam tip: Federation with sync needs two accounts per user, while Workforce Identity Federation needs only the external account.
Continue with this study guide: Managing users and groups in Cloud Identity (manually and automated)
Enabling APIs within projects means turning on a service interface for one specific project before that project can use it. Enabling an API (opens in a new tab) accepts its Terms of Service and billing responsibility for that project, and the same API can stay disabled in another project. Cloud APIs give code a standard JSON REST way to reach services, with client libraries handling connection and authentication.
The API Library is the catalog where APIs are browsed by category or search, showing service name, status, and client library instructions. To enable from the console, select the target project, open the API page, and click Enable. To enable from the CLI, point gcloud at the correct project with billing verified, then run gcloud services enable with service names such as pubsub.googleapis.com. Enabling one API automatically enables the APIs it depends on, but gcloud services list --project $PROJECT_ID should still be used to verify every required name is present.
API activation requires the serviceusage.services.enable permission, supplied by Owner or Service Usage Admin, while Project Creator alone only permits creating a project. Billing must be linked first because usage is charged to the billing account associated with the project where the API is enabled. When a call touches multiple projects, the client project owns the credentials and the resource project owns the target, and client-based APIs charge the client project while resource-based APIs charge the resource project.
Exam tip: APIs are enabled per project, not for an entire organization, and enabling requires serviceusage.services.enable, supplied by Service Usage Admin when the caller is not a project Owner.
Continue with this study guide: Enabling APIs within projects
Provisioning products in Google Cloud Observability means setting up logging, monitoring, and alerting so project health can be seen and investigated. Google Cloud Observability (opens in a new tab) collects telemetry from resources and applications into Cloud Monitoring for metrics and Cloud Logging for log storage and search. The Observability API must be enabled in a billed project before managing buckets, views, and links.
The Ops Agent is a single agent on Compute Engine virtual machine (VM) instances that sends metrics to Cloud Monitoring and logs to Cloud Logging. Telemetry flows from the Ops Agent on each VM to those services, where it appears in charts, dashboards, and Logs Explorer. Each VM needs the cloud-platform access scope and a service account with Monitoring Metric Writer and Logs Writer, and a VM Extension Manager policy installs and maintains the agent across a fleet selected by labels and OS type with version set to current-major and autoupgrade enabled.
A log bucket is the storage container that controls retention for all logs it holds, with custom retention from 1 to 3650 days and default encryption at rest. A log sink is a routing rule that forwards matching logs to BigQuery for SQL analysis, Cloud Storage for low-cost archival, or Pub/Sub for real-time delivery. A custom dashboard visualizes internal metrics such as CPU, memory, network, disk I/O, and response times, while an uptime check probes a public HTTP, HTTPS, or TCP endpoint from Google Cloud checkers and triggers an alerting policy through email, PagerDuty, or Slack when it fails.
Exam tip: A custom retention rule covers an entire log bucket for 1 to 3650 days; it cannot set different retention for individual log types inside the same bucket.
Continue with this study guide: Provisioning and setting up products in Google Cloud Observability
Assessing quotas and requesting increases means comparing current use against the allowed maximum and formally asking to raise an adjustable ceiling. Quotas (opens in a new tab) generally apply per Google Cloud project and restrict consumption to ensure fairness and prevent overload, while system limits are fixed attributes such as identifier length that cannot be changed. Rate limits control how often an action can happen over time, such as API calls per minute, while allocation limits control how many resources can exist at once, such as load balancers or projects.
The Quotas page shows current use against each quota and links to the adjustment form for that service, metric, and region. Checking quotas requires Project Owner, Project Editor, or Quota Viewer, while modifying quotas requires Owner, Editor, Quota Administrator, or a custom role with serviceusage.quotas.update. The request workflow selects the quotas, clicks Edit quotas, enters contact details and justification with the new value, submits for review, and waits 24 to 48 hours before scaling.
Quota scope defines the boundary a limit applies to, and both regional and project-wide quotas must have room before creation succeeds. For example, Cloud Run allows 1,000 services and 1,000 jobs per project and region, and an increase in one region does not raise another region. Available quota does not guarantee physical capacity, because a zone can run out of a resource such as IP addresses even when quota remains, in which case the path is to use another project or region.
Exam tip: Available quota does not guarantee resource availability — a zone can run out of a resource even when the project still has quota left.
Continue with this study guide: Assessing quotas and requesting increases
Setting up standalone organizations means creating the root Organization resource that gives a company central ownership of all folders, projects, and billing accounts. The Organization resource is automatically provisioned after domain verification through Google Workspace or Cloud Identity, with no separate create command. Each identity account links to exactly one organization tied to exactly one domain, and projects created before the organization stay under No organization until moved in.
The Super administrator is the domain-level administrator who verifies the domain, creates the organization, and assigns the Organization Administrator role, then stays available for recovery. The Organization Administrator (roles/resourcemanager.organizationAdmin) has full control over one organization and its IAM policies but cannot access other organizations and by default cannot create folders or projects without extra roles. Cloud Billing (opens in a new tab) administration is kept separate, with Billing Account Creator bringing accounts into the organization, Billing Account Administrator managing payment and links, and Billing Account User linking projects to pay for them.
The hierarchy places the Organization at the root, optional Folders in the middle, and Projects at the base to hold service resources, with every child having exactly one parent. Folders can nest up to 10 levels deep to model departments or environments such as production and non-production. The effective policy at any node equals its directly attached policies plus all policies inherited from its ancestors, so one grant at the top applies consistently below.
Exam tip: The Billing Account Administrator manages an existing billing account, while the Billing Account Creator is needed to create billing accounts and bring existing accounts into the organization.
Continue with this study guide: Setting up standalone organizations
Setting up cloud networking means creating private networks, address space, firewall rules, and name resolution for Google Cloud workloads. A Virtual Private Cloud (VPC) (opens in a new tab) network is the top-level private container that holds resources and controls how they connect, while each subnet inside it belongs to one region and supplies IP addresses in that region. Every workload needing private connectivity depends on a VPC network and a subnet for its address.
Traffic flows from regional resources to endpoint IPs allocated from subnets in the same region as the service attachment, with IPv4 addresses allocated automatically unless a static reservation is used and IPv6 requiring a static /96 reservation. A static internal address stays reserved after its VM is deleted, while an ephemeral address is released back to the subnet pool. Shared VPC keeps the VPC network and subnets in a host project while workloads live in service projects, requiring Compute Network Admin on the service project and Compute Network User on the host project.
VPC firewall rules are defined at the VPC level but enforced on each VM, and they can target all instances, instances with specific network tags, or instances with specific service accounts. A custom network denies all ingress by default and needs a higher-priority allow rule for VM-to-VM traffic, while the default network allows internal traffic through default-allow-internal. Cloud DNS translates names to IPs, with private zones visible only to authorized VPC networks and public zones publishing authoritative records for the internet.
Exam tip: A custom network blocks ingress until you create a higher-priority ingress allow rule, while the default network already allows internal VM-to-VM traffic.
Continue with this study guide: Setting up cloud networking
Confirming availability in geographical locations means verifying that a service, feature, or capacity exists in a target region or zone before deployment. The three primary verification tools are the Google Cloud Console for visual checks, the gcloud CLI for terminal queries, and the official Cloud Locations page for product-by-location lists. Verification must account for resource scope, because global, regional, and zonal resources follow different reachability rules.
Regions and zones (opens in a new tab) define where resources can be used and what failures they survive. A global resource is reachable from any region or zone, a regional resource is usable in any zone of the same region, and a zonal resource is usable only in the same zone. For example, VPC networks and images are global, subnets and regional disks are regional, and Compute Engine VM instances and zonal disks are zonal, so a disk can only attach to a VM in the same zone.
Wider scope generally brings higher availability: single-zone deployments target 99.9%, multi-zone regional deployments target 99.99%, and multi-region deployments target 99.999%. Traffic flows from clients to a global load balancer, then to regional backends in multiple zones, then to replicated data with backups in another location for region failover. Colocating dependent resources such as Compute Engine and Cloud Storage in the same region reduces latency and avoids cross-region egress, while VPC Service Controls perimeters keep data and API calls inside approved networks.
Exam tip: Disks can only be attached to computers in the same zone, so verify both the VM and the disk in that exact zone.
Configuring Cloud Asset Inventory and using Gemini Cloud Assist to analyze resources means keeping a searchable metadata record and asking plain-language questions about it. Cloud Asset Inventory is a global metadata service that collects resource, policy, and relationship data for compliance tracking, audits, and analytics. Gemini Cloud Assist (opens in a new tab) is a generative AI assistant in the Cloud Assist panel that reads telemetry, logs, and inventory metadata to answer questions, suggest code, and recommend fixes.
An asset export is a point-in-time snapshot that flows from Cloud Asset Inventory to Cloud Storage for audit files or to BigQuery for SQL analytics, preserving history beyond the 35-day change window. Search flows from the caller with a project, folder, or organization scope, using searchAllResources for metadata and searchAllIamPolicies for allow and deny policies, with snapshot time within the last 35 days and up to 500 results per page. Effective-permission analysis with analyze-iam-policy combines inherited bindings and custom roles, requiring permissions to read both resources and policies plus iam.roles.get.
Gemini Cloud Assist investigations combine logs and metrics into Issue, Relevant Observations, and Hypotheses with recommended fixes that can be attached to a support case. Cloud Asset Inventory uses periodic sync with eventual consistency, so it fits scheduled audits, drift reporting, and delay-tolerant Pub/Sub notifications. The service API directly fits immediate validation, such as verifying a firewall rule before cutover or revoking access right now, where stale inventory data could break availability or security.
Exam tip: Use Cloud Asset Inventory for scheduled audits and delay-tolerant reconciliation, but use the service API directly when you need immediate validation before a cutover.
serviceusage.services.enable, and enabling one API automatically enables the APIs it depends on.Identity and Access Management (IAM) controls who can act on resources, while Organization Policy controls what configuration is allowed on resources. IAM roles, tags, or Organization Policy rules attached at the organization or folder are inherited automatically by each child, and Google Cloud blocks creation or update of a resource that violates an enforced constraint.
A policy enforced on an organization covers all folders and projects under it, and a policy enforced on a folder covers all projects in that folder including projects added later. An exception requires a different policy set directly on the lower-level folder or project, so only that branch operates under the exception while the strict baseline remains elsewhere.
For least-privilege access, use a predefined role defined by Google for a specific service or job or a custom role defined by the customer organization for an exact permission set, and do not use primitive Owner, Editor, or Viewer roles in production because they grant far more access than most tasks need. A role granted on a project, folder, or organization applies through an allow policy that connects a principal, a role, and a resource and inherits downward to all contained resources.
Cloud APIs must be enabled per project, not for an entire organization, and enabling one API automatically enables the APIs it depends on. Enabling an API accepts its Terms of Service and billing responsibility for that project, requires the serviceusage.services.enable permission supplied by Owner or Service Usage Admin, and requires billing to be linked first because usage is charged to the billing account associated with the project where the API is enabled.
Last updated