multiple instances do not provide zone-level resilience if all run in one zone. In this case, two application VMs must continue after the loss of one availability zone. “A single availability-zone outage can affect both VMs” would not satisfy that requirement; the supported action is “Deploy instances across separate supported availability zones”. Zone placement protects against a zone-level failure when the application is designed for redundancy.

zone-distributed backends plus a load balancer improve resilience and provide a consistent entry point. In this case, two application VMs must continue after the loss of one availability zone. “Zone-spread VM instances with a suitable load-balancing layer” would not satisfy that requirement; the supported action is “Deploy instances across separate supported availability zones”. Zone placement protects against a zone-level failure when the application is designed for redundancy.

availability sets spread eligible VMs across fault and update domains within a region. In this case, two application VMs must continue after the loss of one availability zone. “An availability set with appropriate fault-domain distribution” would not satisfy that requirement; the supported action is “Deploy instances across separate supported availability zones”. Zone placement protects against a zone-level failure when the application is designed for redundancy.

availability sets spread eligible VMs across fault and update domains within a region. In this case, a web tier must add or remove identical VM instances as CPU demand changes. “An availability set with appropriate fault-domain distribution” would not satisfy that requirement; the supported action is “A Virtual Machine Scale Set with autoscale rules”. VM scale sets manage a fleet of similar instances and can scale membership from metrics.

scale-set health integration helps detect and replace unhealthy instances rather than relying only on raw capacity. In this case, a web tier must add or remove identical VM instances as CPU demand changes. “Health monitoring and appropriate instance replacement or repair settings” would not satisfy that requirement; the supported action is “A Virtual Machine Scale Set with autoscale rules”. VM scale sets manage a fleet of similar instances and can scale membership from metrics.

policy-driven scaling limits overprovisioning while retaining enough capacity during peak windows. In this case, a web tier must add or remove identical VM instances as CPU demand changes. “Autoscale thresholds and minimum/maximum instance counts appropriate to demand” would not satisfy that requirement; the supported action is “A Virtual Machine Scale Set with autoscale rules”. VM scale sets manage a fleet of similar instances and can scale membership from metrics.

zone placement protects against a zone-level failure when the application is designed for redundancy. In this case, a team plans to place two highly available VMs in the same zone only. “Deploy instances across separate supported availability zones” would not satisfy that requirement; the supported action is “A single availability-zone outage can affect both VMs”. Multiple instances do not provide zone-level resilience if all run in one zone.

zone-distributed backends plus a load balancer improve resilience and provide a consistent entry point. In this case, a team plans to place two highly available VMs in the same zone only. “Zone-spread VM instances with a suitable load-balancing layer” would not satisfy that requirement; the supported action is “A single availability-zone outage can affect both VMs”. Multiple instances do not provide zone-level resilience if all run in one zone.

vM scale sets manage a fleet of similar instances and can scale membership from metrics. In this case, a team plans to place two highly available VMs in the same zone only. “A Virtual Machine Scale Set with autoscale rules” would not satisfy that requirement; the supported action is “A single availability-zone outage can affect both VMs”. Multiple instances do not provide zone-level resilience if all run in one zone.

policy-driven scaling limits overprovisioning while retaining enough capacity during peak windows. In this case, a VM scale set must keep operating when some instances fail health checks. “Autoscale thresholds and minimum/maximum instance counts appropriate to demand” would not satisfy that requirement; the supported action is “Health monitoring and appropriate instance replacement or repair settings”. Scale-set health integration helps detect and replace unhealthy instances rather than relying only on raw capacity.

vM scale sets manage a fleet of similar instances and can scale membership from metrics. In this case, a VM scale set must keep operating when some instances fail health checks. “A Virtual Machine Scale Set with autoscale rules” would not satisfy that requirement; the supported action is “Health monitoring and appropriate instance replacement or repair settings”. Scale-set health integration helps detect and replace unhealthy instances rather than relying only on raw capacity.

multiple instances do not provide zone-level resilience if all run in one zone. In this case, a VM scale set must keep operating when some instances fail health checks. “A single availability-zone outage can affect both VMs” would not satisfy that requirement; the supported action is “Health monitoring and appropriate instance replacement or repair settings”. Scale-set health integration helps detect and replace unhealthy instances rather than relying only on raw capacity.

zone placement protects against a zone-level failure when the application is designed for redundancy. In this case, an application needs capacity across several zones and a stable front-end endpoint. “Deploy instances across separate supported availability zones” would not satisfy that requirement; the supported action is “Zone-spread VM instances with a suitable load-balancing layer”. Zone-distributed backends plus a load balancer improve resilience and provide a consistent entry point.

multiple instances do not provide zone-level resilience if all run in one zone. In this case, an application needs capacity across several zones and a stable front-end endpoint. “A single availability-zone outage can affect both VMs” would not satisfy that requirement; the supported action is “Zone-spread VM instances with a suitable load-balancing layer”. Zone-distributed backends plus a load balancer improve resilience and provide a consistent entry point.

scale-set health integration helps detect and replace unhealthy instances rather than relying only on raw capacity. In this case, an application needs capacity across several zones and a stable front-end endpoint. “Health monitoring and appropriate instance replacement or repair settings” would not satisfy that requirement; the supported action is “Zone-spread VM instances with a suitable load-balancing layer”. Zone-distributed backends plus a load balancer improve resilience and provide a consistent entry point.

scale-set health integration helps detect and replace unhealthy instances rather than relying only on raw capacity. In this case, a scale set is increasing cost even though peak traffic occurs only at noon. “Health monitoring and appropriate instance replacement or repair settings” would not satisfy that requirement; the supported action is “Autoscale thresholds and minimum/maximum instance counts appropriate to demand”. Policy-driven scaling limits overprovisioning while retaining enough capacity during peak windows.

availability sets spread eligible VMs across fault and update domains within a region. In this case, a scale set is increasing cost even though peak traffic occurs only at noon. “An availability set with appropriate fault-domain distribution” would not satisfy that requirement; the supported action is “Autoscale thresholds and minimum/maximum instance counts appropriate to demand”. Policy-driven scaling limits overprovisioning while retaining enough capacity during peak windows.

zone placement protects against a zone-level failure when the application is designed for redundancy. In this case, a scale set is increasing cost even though peak traffic occurs only at noon. “Deploy instances across separate supported availability zones” would not satisfy that requirement; the supported action is “Autoscale thresholds and minimum/maximum instance counts appropriate to demand”. Policy-driven scaling limits overprovisioning while retaining enough capacity during peak windows.

availability sets spread eligible VMs across fault and update domains within a region. In this case, a production VM requires a restart for resizing but the team needs predictable disruption. “An availability set with appropriate fault-domain distribution” would not satisfy that requirement; the supported action is “A maintenance window and any required VM deallocation before resizing”. Some size changes interrupt the running instance, so the team must plan operational impact.

zone placement protects against a zone-level failure when the application is designed for redundancy. In this case, a production VM requires a restart for resizing but the team needs predictable disruption. “Deploy instances across separate supported availability zones” would not satisfy that requirement; the supported action is “A maintenance window and any required VM deallocation before resizing”. Some size changes interrupt the running instance, so the team must plan operational impact.

vM scale sets manage a fleet of similar instances and can scale membership from metrics. In this case, a production VM requires a restart for resizing but the team needs predictable disruption. “A Virtual Machine Scale Set with autoscale rules” would not satisfy that requirement; the supported action is “A maintenance window and any required VM deallocation before resizing”. Some size changes interrupt the running instance, so the team must plan operational impact.

Raw blob containers do not provide a managed OCI registry repository interface. In this case, a company wants a private repository for images used by Azure container deployments. “Store OCI images as ordinary blobs without a registry endpoint” would not satisfy that requirement; the supported action is “Azure Container Registry (ACR)”. ACR stores and distributes container images under Azure identity and network controls.

ACI executes containers and is not the container image repository. In this case, a company wants a private repository for images used by Azure container deployments. “Deploy an Azure Container Instance and treat it as the image registry” would not satisfy that requirement; the supported action is “Azure Container Registry (ACR)”. ACR stores and distributes container images under Azure identity and network controls.

Private DNS resolves names but does not store and distribute container images. In this case, a company wants a private repository for images used by Azure container deployments. “Configure a private Azure DNS zone alone” would not satisfy that requirement; the supported action is “Azure Container Registry (ACR)”. ACR stores and distributes container images under Azure identity and network controls.

digest references select exact image content instead of a potentially retargeted tag. In this case, an ACR image tag was pushed but production still deploys an older digest. “A verified image digest for immutable artifact selection” would not satisfy that requirement; the supported action is “The image repository, tag-to-digest mapping and deployment reference”. Container deployments resolve tags and digests; an incorrect reference can keep running older content.

private image downloads require the runtime principal to have pull permission. In this case, an ACR image tag was pushed but production still deploys an older digest. “Registry pull authorization for the deployment identity” would not satisfy that requirement; the supported action is “The image repository, tag-to-digest mapping and deployment reference”. Container deployments resolve tags and digests; an incorrect reference can keep running older content.

image pushes address the registry endpoint and repository path for the artifact. In this case, an ACR image tag was pushed but production still deploys an older digest. “The intended Azure Container Registry login server and repository” would not satisfy that requirement; the supported action is “The image repository, tag-to-digest mapping and deployment reference”. Container deployments resolve tags and digests; an incorrect reference can keep running older content.

image pushes address the registry endpoint and repository path for the artifact. In this case, a build agent needs permission to push images but not administer the registry. “The intended Azure Container Registry login server and repository” would not satisfy that requirement; the supported action is “The appropriate registry/repository push permission under its RBAC mode”. Push permissions authorize artifact upload without needing broad registry management privileges.

identity-based registry authorization avoids embedding broadly reusable credentials. In this case, a build agent needs permission to push images but not administer the registry. “Managed identity with scoped registry access where supported” would not satisfy that requirement; the supported action is “The appropriate registry/repository push permission under its RBAC mode”. Push permissions authorize artifact upload without needing broad registry management privileges.

aCR stores and distributes container images under Azure identity and network controls. In this case, a build agent needs permission to push images but not administer the registry. “Azure Container Registry (ACR)” would not satisfy that requirement; the supported action is “The appropriate registry/repository push permission under its RBAC mode”. Push permissions authorize artifact upload without needing broad registry management privileges.

identity-based registry authorization avoids embedding broadly reusable credentials. In this case, a container deployment fails because it cannot pull a private registry image. “Managed identity with scoped registry access where supported” would not satisfy that requirement; the supported action is “Registry pull authorization for the deployment identity”. Private image downloads require the runtime principal to have pull permission.

image pushes address the registry endpoint and repository path for the artifact. In this case, a container deployment fails because it cannot pull a private registry image. “The intended Azure Container Registry login server and repository” would not satisfy that requirement; the supported action is “Registry pull authorization for the deployment identity”. Private image downloads require the runtime principal to have pull permission.

aCR stores and distributes container images under Azure identity and network controls. In this case, a container deployment fails because it cannot pull a private registry image. “Azure Container Registry (ACR)” would not satisfy that requirement; the supported action is “Registry pull authorization for the deployment identity”. Private image downloads require the runtime principal to have pull permission.