Try each one before opening the answer. Every option is explained, with the Google Cloud documentation page that proves it.
Question 1 · choose 1
Every GKE cluster created or updated in the prod folder must have Binary Authorization turned on. A request that would create or update a cluster without it must fail, not just be reported after the fact, and the security team does not want to write or maintain condition expressions of its own for this rule. What should the security engineer implement?
AA log-based alert on cluster creation and update audit entries that lack Binary Authorization
BA custom organization policy constraint on GKE clusters that denies requests where Binary Authorization is off
CAn organization policy on the prod folder that enforces the managed constraint container.managed.enableBinaryAuthorization
DA Binary Authorization policy in each prod project that requires attestations for every image
Show the answer and why
AA log-based alert on cluster creation and update audit entries that lack Binary Authorization
Incorrect
An alert fires after the cluster has been created or changed, which the team said is not enough, and its query is logic the team would maintain.
BA custom organization policy constraint on GKE clusters that denies requests where Binary Authorization is off
Incorrect
It would block the request, but a custom constraint's condition is a CEL expression the team must write and maintain, and Google already publishes a managed constraint for this setting.
CAn organization policy on the prod folder that enforces the managed constraint container.managed.enableBinaryAuthorization
Correct
This managed constraint requires Binary Authorization to be enabled when GKE clusters are created or updated, so a noncompliant request fails, and Google maintains its logic.
DA Binary Authorization policy in each prod project that requires attestations for every image
Incorrect
The policy governs image deployment on clusters where Binary Authorization is enabled; it does not stop anyone from creating a cluster with it turned off.
Google publishes managed constraints for many GKE settings, among them container.managed.enableBinaryAuthorization ("Require enabling Binary Authorization for GKE clusters"), which applies when clusters are created or updated. Enforced through a policy on the folder, it makes a noncompliant request fail. A custom constraint is the tool for a setting that no Google-provided constraint covers, and it needs a CEL condition of your own; an alert only reports after the fact.
The organization policy for resource locations allows only in:eu-locations. One research project must also create resources in us-central1, while still being able to use every EU location it uses today. No other project may change. How should the security engineer set the policy on that project?
AAllow us-central1 locations in a project policy that merges with the parent's allowed values
BAllow us-central1 locations in a project policy that replaces the parent policy
CAdd us-central1 locations to the organization's policy
DRemove the policy from the project's folder so that the project falls back to the default
Show the answer and why
AAllow us-central1 locations in a project policy that merges with the parent's allowed values
Correct
With inheritFromParent set to true, the project's allowed values are merged with the inherited ones, so the project gets the EU locations plus us-central1.
BAllow us-central1 locations in a project policy that replaces the parent policy
Incorrect
A policy that does not inherit replaces the parent's values, so the project would lose the EU locations it uses today.
CAdd us-central1 locations to the organization's policy
Incorrect
Changing the organization policy would let every project use us-central1, which the requirement forbids.
DRemove the policy from the project's folder so that the project falls back to the default
Incorrect
Removing a folder policy changes every project in the folder, and the project would still inherit the organization's EU-only rule.
Organization policies are inherited down the hierarchy. A policy on a lower resource either replaces the inherited one or, with inheritFromParent set, is merged and reconciled with it. For list constraints such as resource locations, merging adds the child's allowed values to the parent's.
An organization created in 2025 cannot grant a viewer role to users of a partner company's Google Workspace domain; the console rejects the grant with an organization policy error. No one in the company has changed any organization policy. What should the security engineer do to allow the partner?
AGrant the role at the folder level instead, because the policy applies only to projects
BTurn off the IAM recommender, which blocks grants to identities outside the domain
CAdd the partner's Google Workspace customer ID to the allowed values of the domain-restricted sharing policy
DAsk the partner to create Gmail accounts, which the default policy allows
Show the answer and why
AGrant the role at the folder level instead, because the policy applies only to projects
Incorrect
Organization policies are inherited by folders and projects alike, so the grant would fail there too.
BTurn off the IAM recommender, which blocks grants to identities outside the domain
Incorrect
The recommender only suggests changes; it never blocks a grant.
CAdd the partner's Google Workspace customer ID to the allowed values of the domain-restricted sharing policy
Correct
Organizations created on or after May 3, 2024 enforce iam.allowedPolicyMemberDomains by default with only their own domain, so the partner has to be added to the allowed list.
DAsk the partner to create Gmail accounts, which the default policy allows
Incorrect
The default policy allows only the company's own domain, so private Gmail accounts are rejected as well.
Domain-restricted sharing limits which principals can be granted IAM roles in an organization. For organizations created on or after May 3, 2024, the legacy constraint iam.allowedPolicyMemberDomains is enforced by default with the organization's own domain as the only allowed value; the newer iam.managed.allowedPolicyMembers constraint offers the same control with principal sets.
The security team plans to enforce the managed constraint compute.managed.vmExternalIpAccess, which blocks external IP addresses on VMs, across the organization, but fears it will break workloads nobody has listed. It wants to see which create and update requests would be refused, without refusing any of them yet. What should the security engineer do?
APut the policy in dry-run mode and review the audit logs it writes
BEnforce the policy on one production project for a week and watch for incidents
CRun Policy Simulator for allow policies against the new constraint
DTurn on VPC Flow Logs everywhere to find VMs that receive internet traffic today
Show the answer and why
APut the policy in dry-run mode and review the audit logs it writes
Correct
A dry-run policy is evaluated like an enforced one, but violations are only written to audit logs instead of being denied.
BEnforce the policy on one production project for a week and watch for incidents
Incorrect
Enforcing it refuses requests in that project, which is exactly the breakage the team wants to avoid.
CRun Policy Simulator for allow policies against the new constraint
Incorrect
Policy Simulator previews changes to IAM allow policies, not organization policy constraints.
DTurn on VPC Flow Logs everywhere to find VMs that receive internet traffic today
Incorrect
Flow logs show current traffic but not which future API requests the constraint would refuse.
Organization policies can be created in dry-run mode. They are evaluated like enforced policies, but a violating action is allowed and recorded in the audit logs, so a team can measure the impact of a constraint before turning on enforcement. Dry run is available for managed and custom constraints and a few legacy ones, which is why the managed constraint is used here rather than the legacy compute.vmExternalIpAccess.
Production and test projects are mixed inside many folders. A constraint that blocks service account key creation must apply to every project that carries the secure tag environment set to production, and to no other project, including projects tagged later. How should the security engineer apply the constraint?
AReorganize all folders so that production projects sit in one folder, then set the policy on that folder
BOne organization-level policy whose rule is enforced only under a condition on that tag
CA separate policy on each production project, set when the project is created
DAn IAM deny policy on the key creation permission, attached to every production project
Show the answer and why
AReorganize all folders so that production projects sit in one folder, then set the policy on that folder
Incorrect
Moving projects could work but is a large change with its own risk, and it is not needed for this requirement.
BOne organization-level policy whose rule is enforced only under a condition on that tag
Correct
Organization policies can be enforced conditionally on secure tags, so the rule follows the tag wherever a project sits.
CA separate policy on each production project, set when the project is created
Incorrect
Per-project policies depend on every creator and miss projects that are tagged as production later.
DAn IAM deny policy on the key creation permission, attached to every production project
Incorrect
Attached project by project, the deny policy misses projects tagged as production later, and the requirement is for the constraint itself to follow the tag.
Tags can scope organization policies: a policy rule can carry a condition on a tag key and value, and it is enforced only on resources that have, or inherit, that tag. This lets one policy target production resources wherever they sit in the hierarchy.
The Shared VPC host project carries networking for 200 service projects. An engineer nearly deleted it by mistake during a cleanup. The network team wants a safeguard that makes any delete request for this one project fail until someone deliberately removes the safeguard. What should the security engineer apply?
ARemove the Owner role from the network team so that nobody holds the project delete permission
BA daily export of the project's configuration so that it can be rebuilt after a deletion
CAn organization policy that disables the Resource Manager API in the host project
DA lien on the host project that restricts resourcemanager.projects.delete until it is removed
Show the answer and why
ARemove the Owner role from the network team so that nobody holds the project delete permission
Incorrect
Taking away Owner also removes permissions the team needs, and someone must still hold the permission.
BA daily export of the project's configuration so that it can be rebuilt after a deletion
Incorrect
A backup helps recovery after the damage; it does not block the deletion.
CAn organization policy that disables the Resource Manager API in the host project
Incorrect
Disabling the API in the project would break project management and is not how deletion is protected.
DA lien on the host project that restricts resourcemanager.projects.delete until it is removed
Correct
A lien blocks deletion of the project until an Organization Administrator or Project Owner removes it.
Project liens protect critical projects from accidental or malicious deletion. While a lien exists, the project cannot be deleted, and Google Cloud also places liens automatically in some cases, such as on projects whose service accounts are used across projects.