Question 20
The platform operations team must manage bursty training demand without uncontrolled costs. The team is required to cap simultaneous training capacity to a financially approved ceiling. Which option is the best fit?
- Minimum node count zero
- Maximum node count
- Regional quota request
- Job concurrency limit
Explanation
Correct Answer
B:The maximum node setting bounds cluster growth regardless of queued work. It is a capacity guardrail rather than a guarantee that every queued job starts immediately.
Incorrect Answers
A:Setting the minimum node count to zero lets a compatible cluster deallocate idle nodes; running jobs can still trigger provisioning, which introduces startup latency; the outstanding need is to cap simultaneous training capacity to a financially approved ceiling.
C:Azure VM family quotas constrain compute provisioning even when ML job YAML is otherwise valid; capacity or quota must be addressed before the requested nodes can be allocated; the needed capability is to cap simultaneous training capacity to a financially approved ceiling.
D:A concurrency setting controls parallel task execution and pressure on downstream resources; it is distinct from the total number of compute nodes available to the workspace; the requirement, by contrast, is to cap simultaneous training capacity to a financially approved ceiling.
Question 21
The platform operations team must manage bursty training demand without uncontrolled costs. An operator is tasked to shorten the time a compute node remains allocated after finishing its last task. Which mechanism provides that capability?
- Maximum node count
- Idle scale-down time
- Regional quota request
- Low-priority compute selection
- Job concurrency limit
Explanation
Correct Answer
B:The idle time policy governs scale-down delay for managed cluster nodes. It influences cost and startup churn without changing how a model's metrics are evaluated.
Incorrect Answers
A:The maximum node setting bounds cluster growth regardless of queued work; it is a capacity guardrail rather than a guarantee that every queued job starts immediately; the requirement, by contrast, is to shorten the time a compute node remains allocated after finishing its last task.
C:Azure VM family quotas constrain compute provisioning even when ML job YAML is otherwise valid; capacity or quota must be addressed before the requested nodes can be allocated; the specified outcome instead requires the team to shorten the time a compute node remains allocated after finishing its last task.
D:Low-priority nodes trade stability for lower compute cost and may be preempted; they are inappropriate for a job that must never be interrupted without recovery logic; the implementation under review needs to shorten the time a compute node remains allocated after finishing its last task.
E:A concurrency setting controls parallel task execution and pressure on downstream resources; it is distinct from the total number of compute nodes available to the workspace; the requested task remains to shorten the time a compute node remains allocated after finishing its last task.
Question 22
The platform operations team must manage bursty training demand without uncontrolled costs. The immediate design goal is to resolve a training job that cannot allocate its required VM family because subscription quota is exhausted. Which approach addresses that goal?
- Job concurrency limit
- Regional quota request
- Low-priority compute selection
- Maximum node count
Explanation
Correct Answer
B:Azure VM family quotas constrain compute provisioning even when ML job YAML is otherwise valid. Capacity or quota must be addressed before the requested nodes can be allocated.
Incorrect Answers
A:A concurrency setting controls parallel task execution and pressure on downstream resources; it is distinct from the total number of compute nodes available to the workspace; the implementation under review needs to resolve a training job that cannot allocate its required VM family because subscription quota is exhausted.
C:Low-priority nodes trade stability for lower compute cost and may be preempted; they are inappropriate for a job that must never be interrupted without recovery logic; this alternative does not achieve the requirement to resolve a training job that cannot allocate its required VM family because subscription quota is exhausted.
D:The maximum node setting bounds cluster growth regardless of queued work; it is a capacity guardrail rather than a guarantee that every queued job starts immediately; the actual design criterion is to resolve a training job that cannot allocate its required VM family because subscription quota is exhausted.
Question 23
The platform operations team must manage bursty training demand without uncontrolled costs. The team is required to reduce costs for interruption-tolerant experiments that can retry when spare capacity disappears. Which option is the best fit?
- Low-priority compute selection
- Maximum node count
- Minimum node count zero
- Idle scale-down time
Explanation
Correct Answer
A:Low-priority nodes trade stability for lower compute cost and may be preempted. They are inappropriate for a job that must never be interrupted without recovery logic.
Incorrect Answers
B:The maximum node setting bounds cluster growth regardless of queued work; it is a capacity guardrail rather than a guarantee that every queued job starts immediately; the outstanding need is to reduce costs for interruption-tolerant experiments that can retry when spare capacity disappears.
C:Setting the minimum node count to zero lets a compatible cluster deallocate idle nodes; running jobs can still trigger provisioning, which introduces startup latency; this approach leaves the team unable to reduce costs for interruption-tolerant experiments that can retry when spare capacity disappears.
D:The idle time policy governs scale-down delay for managed cluster nodes; it influences cost and startup churn without changing how a model's metrics are evaluated; the needed capability is to reduce costs for interruption-tolerant experiments that can retry when spare capacity disappears.
Question 24
The platform operations team must manage bursty training demand without uncontrolled costs. An operator is tasked to restrict how many independent tasks a scheduled workload runs at the same time. Which mechanism provides that capability?
- Low-priority compute selection
- Idle scale-down time
- Maximum node count
- Job concurrency limit
Explanation
Correct Answer
D:A concurrency setting controls parallel task execution and pressure on downstream resources. It is distinct from the total number of compute nodes available to the workspace.
Incorrect Answers
A:Low-priority nodes trade stability for lower compute cost and may be preempted; they are inappropriate for a job that must never be interrupted without recovery logic; the needed capability is to restrict how many independent tasks a scheduled workload runs at the same time.
B:The idle time policy governs scale-down delay for managed cluster nodes; it influences cost and startup churn without changing how a model's metrics are evaluated; the requirement, by contrast, is to restrict how many independent tasks a scheduled workload runs at the same time.
C:The maximum node setting bounds cluster growth regardless of queued work; it is a capacity guardrail rather than a guarantee that every queued job starts immediately; the team would still need to restrict how many independent tasks a scheduled workload runs at the same time.
Question 25
The security team audits Azure Machine Learning workspace access and separation of duties. The immediate design goal is to allow auditors to inspect workspace configuration and assets without granting edit rights. Which approach addresses that goal?
- Azure ML Data Scientist role
- Azure ML workspace Reader
- Least-privilege review
- Azure RBAC role assignment scope
Explanation
Correct Answer
B:Reader permissions support observation at the assigned resource scope. They do not permit mutation of workspaces, creation of jobs, or management of RBAC assignments.
Incorrect Answers
A:The AzureML Data Scientist built-in role is designed for many ML workspace activities, including experiments and jobs; it is not equivalent to broad subscription Owner rights; the specified outcome instead requires the team to allow auditors to inspect workspace configuration and assets without granting edit rights.
C:Least privilege assigns only actions and scopes needed for operation; broad Owner-style permissions create unnecessary ability to modify unrelated resources and access policies; the requested task remains to allow auditors to inspect workspace configuration and assets without granting edit rights.
D:An assignment at the workspace resource scope limits inherited permissions to that resource and any applicable children; subscription-wide grants would expand the operator's reach; this alternative does not achieve the requirement to allow auditors to inspect workspace configuration and assets without granting edit rights.
Question 26
The security team audits Azure Machine Learning workspace access and separation of duties. The team is required to let a practitioner create and run ML workloads while limiting unrelated Azure infrastructure administration. Which option is the best fit?
- Role-assignment authority
- Azure RBAC role assignment scope
- Azure ML workspace Reader
- Azure ML Data Scientist role
Explanation
Correct Answer
D:The AzureML Data Scientist built-in role is designed for many ML workspace activities, including experiments and jobs. It is not equivalent to broad subscription Owner rights.
Incorrect Answers
A:Managing Azure RBAC assignments requires specific authorization such as an appropriate access administrator role; resource contribution privileges alone should not be assumed to include access delegation; this alternative does not achieve the requirement to let a practitioner create and run ML workloads while limiting unrelated Azure infrastructure administration.
B:An assignment at the workspace resource scope limits inherited permissions to that resource and any applicable children; subscription-wide grants would expand the operator's reach; the actual design criterion is to let a practitioner create and run ML workloads while limiting unrelated Azure infrastructure administration.
C:Reader permissions support observation at the assigned resource scope; they do not permit mutation of workspaces, creation of jobs, or management of RBAC assignments; the outstanding need is to let a practitioner create and run ML workloads while limiting unrelated Azure infrastructure administration.
Question 27
The security team audits Azure Machine Learning workspace access and separation of duties. An operator is tasked to grant an operator access to one workspace rather than every workspace in the subscription. Which mechanism provides that capability?
- Least-privilege review
- Role-assignment authority
- Azure ML Data Scientist role
- Azure RBAC role assignment scope
Explanation
Correct Answer
D:An assignment at the workspace resource scope limits inherited permissions to that resource and any applicable children. Subscription-wide grants would expand the operator's reach.
Incorrect Answers
A:Least privilege assigns only actions and scopes needed for operation; broad Owner-style permissions create unnecessary ability to modify unrelated resources and access policies; this approach leaves the team unable to grant an operator access to one workspace rather than every workspace in the subscription.
B:Managing Azure RBAC assignments requires specific authorization such as an appropriate access administrator role; resource contribution privileges alone should not be assumed to include access delegation; the needed capability is to grant an operator access to one workspace rather than every workspace in the subscription.
C:The AzureML Data Scientist built-in role is designed for many ML workspace activities, including experiments and jobs; it is not equivalent to broad subscription Owner rights; the requirement, by contrast, is to grant an operator access to one workspace rather than every workspace in the subscription.
Question 28
The security team audits Azure Machine Learning workspace access and separation of duties. The immediate design goal is to delegate the ability to create role assignments without assuming ordinary workspace editing grants this permission. Which approach addresses that goal?
- Role-assignment authority
- Least-privilege review
- Azure ML workspace Reader
- Identity authentication check
- Azure RBAC role assignment scope
Explanation
Correct Answer
A:Managing Azure RBAC assignments requires specific authorization such as an appropriate access administrator role. Resource contribution privileges alone should not be assumed to include access delegation.
Incorrect Answers
B:Least privilege assigns only actions and scopes needed for operation; broad Owner-style permissions create unnecessary ability to modify unrelated resources and access policies; the specified outcome instead requires the team to delegate the ability to create role assignments without assuming ordinary workspace editing grants this permission.
C:Reader permissions support observation at the assigned resource scope; they do not permit mutation of workspaces, creation of jobs, or management of RBAC assignments; the implementation under review needs to delegate the ability to create role assignments without assuming ordinary workspace editing grants this permission.
D:Authentication establishes the caller identity and token validity; a well-scoped workspace role cannot compensate for an expired or absent credential; the requested task remains to delegate the ability to create role assignments without assuming ordinary workspace editing grants this permission.
E:An assignment at the workspace resource scope limits inherited permissions to that resource and any applicable children; subscription-wide grants would expand the operator's reach; this alternative does not achieve the requirement to delegate the ability to create role assignments without assuming ordinary workspace editing grants this permission.
Question 29
The security team audits Azure Machine Learning workspace access and separation of duties. The team is required to diagnose a request that has no valid Microsoft Entra token before investigating authorization scopes. Which option is the best fit?
- Azure RBAC role assignment scope
- Identity authentication check
- Azure ML Data Scientist role
- Azure ML workspace Reader
Explanation
Correct Answer
B:Authentication establishes the caller identity and token validity. A well-scoped workspace role cannot compensate for an expired or absent credential.
Incorrect Answers
A:An assignment at the workspace resource scope limits inherited permissions to that resource and any applicable children; subscription-wide grants would expand the operator's reach; the requested task remains to diagnose a request that has no valid Microsoft Entra token before investigating authorization scopes.
C:The AzureML Data Scientist built-in role is designed for many ML workspace activities, including experiments and jobs; it is not equivalent to broad subscription Owner rights; the actual design criterion is to diagnose a request that has no valid Microsoft Entra token before investigating authorization scopes.
D:Reader permissions support observation at the assigned resource scope; they do not permit mutation of workspaces, creation of jobs, or management of RBAC assignments; the outstanding need is to diagnose a request that has no valid Microsoft Entra token before investigating authorization scopes.
Question 30
The security team audits Azure Machine Learning workspace access and separation of duties. An operator is tasked to remove excess permissions from a training service identity after narrowing its actual resource needs. Which mechanism provides that capability?
- Identity authentication check
- Least-privilege review
- Role-assignment authority
- Azure RBAC role assignment scope
Explanation
Correct Answer
B:Least privilege assigns only actions and scopes needed for operation. Broad Owner-style permissions create unnecessary ability to modify unrelated resources and access policies.
Incorrect Answers
A:Authentication establishes the caller identity and token validity; a well-scoped workspace role cannot compensate for an expired or absent credential; the outstanding need is to remove excess permissions from a training service identity after narrowing its actual resource needs.
C:Managing Azure RBAC assignments requires specific authorization such as an appropriate access administrator role; resource contribution privileges alone should not be assumed to include access delegation; the needed capability is to remove excess permissions from a training service identity after narrowing its actual resource needs.
D:An assignment at the workspace resource scope limits inherited permissions to that resource and any applicable children; subscription-wide grants would expand the operator's reach; the requirement, by contrast, is to remove excess permissions from a training service identity after narrowing its actual resource needs.