mcAfee Secure Website
23

Microsoft AZ-104 Bundle

Exam Code: AZ-104

Exam Name Microsoft Azure Administrator

Certification Provider: Microsoft

Corresponding Certification: Microsoft Certified: Azure Administrator Associate

AZ-104 Training Materials $44.99

Reliable & Actual Study Materials for AZ-104 Exam Success

The Latest AZ-104 Exam Questions as Experienced in the Actual Test!

  • 24
    Questions & Answers

    AZ-104 Questions & Answers

    606 Questions & Answers

    Includes questions types found on actual exam such as drag and drop, simulation, type in, and fill in the blank.

  • 25
    AZ-104 Video Course

    AZ-104 Training Course

    132 Video Lectures

    Based on Real Life Scenarios which you will encounter in exam and learn by working with real equipment.

  • exam =30
    Study Guide

    AZ-104 Study Guide

    458 PDF Pages

    Study Guide developed by industry experts who have written exams in the past. They are technology-specific IT certification researchers with at least a decade of experience at Fortune 500 companies.

exam =32

Microsoft Azure Administrator (AZ-104) Practice Tests

Guarantee

Satisfaction Guaranteed

Pass4sure has a remarkable Microsoft Candidate Success record. We're confident of our products and provide no hassle product exchange. That's how confident we are!

99.3% Pass Rate
Total Cost: $194.97
Bundle Price: $149.98

Purchase Individually

  • exam =34
    Questions & Answers

    Questions & Answers

    606 Questions

    $124.99
    exam =35
  • exam =37
    AZ-104 Video Course

    Training Course

    132 Video Lectures

    $39.99
  • exam =36
    Study Guide

    Study Guide

    458 PDF Pages

    $29.99

Microsoft AZ-104 Practice Test 56

Question 550

A Container Apps worker must connect outbound to a database exposed only on an internal Azure network; the database must not become public. What should be configured?

  1. Enable public anonymous ingress on the database to remove network restrictions.
  2. Use an appropriately connected Container Apps environment/VNet routing and private DNS to reach the backend.
  3. Configure an ACR webhook as the only network path to the database.
  4. Use single-revision mode without checking any subnet or private network configuration.

Explanation

Correct Answer

B:

Private VNet connectivity and correct routing/DNS are needed for outbound requests to an internal backend.

Incorrect Answers

A:

Public exposure contradicts the private-backend requirement and increases attack surface.

C:

ACR webhooks notify build events and do not provide database network routing.

D:

Revision mode governs deployment activity, not the private connectivity path.

Question 551

A newly deployed Container Apps revision fails its startup probe and receives no live traffic. What should be investigated before forcing it active?

  1. Delete all healthy old revisions before reading diagnostic data.
  2. Change the public DNS TTL only.
  3. Inspect revision/system logs, probe path/port, container readiness and startup failures.
  4. Increase ACR image storage quota regardless of the startup log messages.

Explanation

Correct Answer

C:

Startup/readiness probes and application logs determine whether a revision is ready to serve traffic.

Incorrect Answers

A:

Removing the healthy fallback before diagnosing the failure risks an unnecessary outage.

B:

DNS TTL cannot make an unhealthy container pass startup probes.

D:

ACR storage quota is not a general remedy for an already-deployed container failing its startup checks.

Question 552

A Container Apps image currently embeds a database password in its filesystem. How should new deployments handle the secret?

  1. Commit the password into an image layer but encrypt the repository name.
  2. Use Container Apps secrets or a supported managed-identity/Key Vault design and reference them at runtime.
  3. Move the password to the image tag so it is not in the filesystem.
  4. Increase the Container Apps maximum replica count to obscure the credential.

Explanation

Correct Answer

B:

Runtime secret references or managed identities avoid baking credential material into immutable container images.

Incorrect Answers

A:

Image layers can be inspected and persisted, so embedding credentials is unsafe even with different registry naming.

C:

Image tags are metadata and are often widely visible; they cannot safely store passwords.

D:

Scaling replica count does not protect secrets embedded in each running image.

Question 553

An Azure Container App has two active revisions in multiple-revision mode. Security now wants to disable external ingress for the entire app, while preserving the existing revision templates and avoiding creation of another revision. Which scope of change meets this requirement?

  1. Change `properties.template.containers[0].image`; the resulting image change updates ingress without creating a revision.
  2. Change the revision-scoped scale rules in `properties.template`; revision-scoped changes apply globally to all existing revisions without creating a new one.
  3. Delete both active revisions and recreate the app; the only way to change ingress is to replace the app.
  4. Update the application-scope ingress configuration under `properties.configuration`; it affects all revisions and does not itself create a new revision.

Explanation

Correct Answer

D:

Ingress settings are application-scope settings in `properties.configuration`. Disabling external ingress applies to every revision without generating a new revision, unlike changes to the image, CPU/memory or scale-rule template.

Incorrect Answers

A:

An image change is revision-scoped because the image belongs to the container template. It produces a new revision; it is not how the application-wide ingress setting is managed.

B:

Scale rules belong to the revision-scoped template. Changing them creates a new revision and does not serve as an application-level ingress toggle for the existing revisions.

C:

Container Apps allows supported ingress changes at application scope without deleting the application or its revisions. Deleting the active revisions would unnecessarily disrupt the running service.

Question 554

Container Apps runs multiple revision mode. A recent release is defective while the previous healthy revision is still retained. How can traffic be recovered quickly?

  1. Disable all revision traffic labels without a replacement destination.
  2. Delete the prior known-good revision before any recovery action.
  3. Shift traffic weights back to the retained healthy previous revision and validate client behavior.
  4. Change the storage account redundancy setting used by the image registry.

Explanation

Correct Answer

C:

Multiple active revisions support redirecting traffic to a previously healthy version for a rapid release rollback.

Incorrect Answers

A:

Removing traffic destinations cannot restore the service to a healthy revision.

B:

Deleting the healthy revision removes the simplest rollback target.

D:

Object-storage redundancy has no direct control over HTTP revision traffic weights.

Question 555

Container App app-catalog exposes HTTP ingress and is intended to incur no running-replica usage charges during long idle periods. It must accept an incoming request after an ordinary cold start. Which TWO scale settings are consistent with scale-to-zero behavior? Select TWO answers.

  1. Use HTTP request concurrency as a supported scale trigger for the ingress-enabled Container App.
  2. Set minReplicas to 0 and choose a supported maximum above zero for its revision.
  3. Configure minReplicas to 2 so a pair of warm replicas always stays available.
  4. Disable ingress and omit every supported scale rule while still expecting HTTP requests to wake the app.
  5. Set maxReplicas to 0 so the application can both wake on demand and process traffic.

Explanation

Correct Answers

A:

An HTTP ingress application can use the built-in HTTP scale rule or a supported custom HTTP-concurrency rule. The scaler monitors incoming requests and can activate replicas from zero, subject to cold-start latency. This provides the demand signal needed to wake an idle app; minimum replicas alone does not define how a non-HTTP workload should activate. Test the expected cold-start behavior and choose a sensible concurrency threshold to avoid underprovisioning during bursts.

B:

Setting the revision minimum replica count to zero permits Azure Container Apps to scale fully down after its scale-in cooldown when there are no requests. A positive supported maximum must remain available for future requests to trigger replicas. With HTTP ingress and an HTTP scaler, this meets the requested idle behavior. Setting minReplicas above zero would keep resources provisioned even with no traffic, defeating the stated ability to scale to zero.

Incorrect Answers

C:

A minimum of two requires two replicas to remain allocated even when no requests arrive. That is the opposite of scale to zero and could incur idle running costs. Maintaining warm capacity might be right for strict low-latency applications, but this case explicitly accepts cold starts to remove idle replicas.

D:

Without ingress or another scale rule, a Container App at zero replicas may have no event source to bring it back. The use case depends on HTTP traffic, so turning off ingress and all triggers undermines that mechanism. Configure a supported HTTP endpoint and request-based scaling rather than making the service unreachable.

E:

Setting the maximum replica count to zero prevents scaling to any running replica; it is not a supported way to provide an on-demand HTTP service. The scale-to-zero requirement concerns the minimum idle capacity, not removal of the maximum usable capacity. Keep maxReplicas at a valid positive value so the HTTP scaler can start work after requests arrive.

Question 556

A Container Apps API must maintain at least two scheduled replicas even when requests stop, and its operators want to prevent a scale-range configuration that caps it below two. The app runs in an environment with sufficient available capacity and quotas. Which TWO revision scale properties must be configured consistently? Select TWO answers.

  1. Set minReplicas to zero so the service can always scale completely down while idle.
  2. Set maxReplicas to at least two so the configured maximum does not contradict the desired warm minimum.
  3. Set maxReplicas to one to ensure that the API never runs more than one replica.
  4. Set minReplicas to two, retaining at least two replicas during normal eligible operation.
  5. Turn off HTTP ingress and all scale rules while expecting ordinary external requests to keep the API reachable.

Explanation

Correct Answers

B:

Container Apps scale limits define the permitted replica-count range. To keep a minimum of two running instances, the maximum must not be smaller than two. Configuring maxReplicas at two or higher leaves room for that baseline and optional additional scaling when traffic increases. Check that the environment can accommodate the selected range; the prompt already states the necessary capacity exists. This upper bound is an independent part of a coherent scale configuration, not a substitute for setting the minimum.

D:

Setting minReplicas to two tells Azure Container Apps that the revision should retain two replicas instead of scaling fully to zero after idle periods. It is the direct scale setting for the required warm baseline. Combine it with maxReplicas of at least two and sufficient environment capacity; operational health still needs monitoring, but health monitoring alone cannot impose a two-replica minimum. During platform upgrades temporary replica counts may vary, so this is a configured target rather than a guarantee against all failures.

Incorrect Answers

A:

A zero minimum explicitly permits the revision to scale to no replicas during idle periods, which conflicts with the requirement to maintain two warm instances. Zero might lower costs for tolerant background or on-demand services, but it does not implement the desired baseline. The service needs a positive minimum equal to the required warm capacity.

C:

A maximum of one prevents the revision from running two replicas even if the minimum setting requests them. This configuration contradicts the case and would create an invalid or unsatisfiable scale range. Choose maxReplicas of at least two alongside the two-replica minimum so that the desired baseline is within the supported limits.

E:

Disabling HTTP ingress would prevent the externally called API from accepting its normal request path, while disabling triggers would not provide an alternative way to scale with demand. Neither setting establishes a warm-replica floor. Keep the supported application access path and configure minReplicas and maxReplicas coherently; unrelated traffic-disabling changes do not satisfy the requirement.

Question 557

An HTTP-enabled Container Apps service reaches its configured maximum replica count while requests queue. Which scale configuration directly controls scale-out capacity?

  1. Increase a private DNS zone TTL.
  2. Decrease the minimum replica count while keeping the same low maximum.
  3. Add another Azure resource-group tag to the app.
  4. Review the HTTP concurrency threshold and increase the supported maxReplicas ceiling as needed.

Explanation

Correct Answer

D:

Container Apps uses HTTP concurrency triggers and min/max replica limits to determine horizontal scaling.

Incorrect Answers

A:

DNS TTL does not alter scaling thresholds or permitted replica count.

B:

Lowering the minimum replica floor cannot raise the maximum service capacity.

C:

Tags organize cloud resources but do not tune runtime scaling.

Question 558

An Azure Container Apps Azure Queue Storage scaler loses authentication after a shared account key rotation. The app has a managed identity. Which robust correction is appropriate?

  1. Increase maximum replicas without restoring queue authentication.
  2. Set the HTTP concurrency target lower even though this is a queue-triggered worker.
  3. Disable all queue permission checks to make KEDA authenticate anonymously.
  4. Configure the scaler for supported managed-identity authentication and grant least-privilege queue access.

Explanation

Correct Answer

D:

Container Apps supports managed-identity authorization for Azure Queue Storage scale rules, avoiding brittle shared account keys.

Incorrect Answers

A:

A scaler cannot inspect backlog when the credential is invalid, regardless of replica ceiling.

B:

HTTP scaling metrics do not fix failed authorization to a Storage Queue event source.

C:

Azure Storage queues require an authorized caller; unauthenticated anonymous triggers are not the safe remedy.

Question 559

A Container Apps replica is OOM-killed because its heap exceeds the memory allocated per replica. What should change?

  1. Increase the per-replica memory allocation within supported limits or reduce its memory footprint.
  2. Increase only maxReplicas while leaving each replica unable to hold the same request.
  3. Change the application’s public DNS label.
  4. Assign AcrPull to a registry whose image has already downloaded successfully.

Explanation

Correct Answer

A:

The memory limit is per replica; increasing its supported capacity or optimizing heap use directly addresses OOM termination.

Incorrect Answers

B:

Additional replicas may spread load but do not inherently fix a single request exceeding each replica’s memory limit.

C:

DNS names do not change process memory allocation.

D:

Registry authorization is not the cause once the image is running and then OOM-killed.

Question 560

A CPU-bound job is currently a single Azure Container Instance, and increasing its CPU capacity is supported. Which next step is directly applicable?

  1. Enable VMSS autoscale settings on the ACI resource itself.
  2. Increase the container group’s CPU allocation within ACI limits, or redesign to an autoscaling service if horizontal elasticity is required.
  3. Change the registry image tag without updating resource requests.
  4. Enable Azure Files soft delete to add CPU cores to the container.

Explanation

Correct Answer

B:

ACI resource requests determine the container group’s CPU capacity; a new configuration/deployment may be needed for more resources.

Incorrect Answers

A:

Standalone ACI does not directly inherit VM Scale Set autoscale policy.

C:

Retagging an image does not increase allocated runtime CPU capacity.

D:

Storage recovery settings cannot increase available compute cores.