Exam Bundle

The active context, API endpoint, and changed resource match the intended environment in the audit trail. That observed result would support another specific correction: Inspect the active kubeconfig context and explicitly select the authorized cluster before applying changes. The other scenario concerns: An operator runs kubectl without selecting the intended cluster context and modifies a development workload by mistake. The current Timberline Foods problem is: An operator rolls back a broken release by deleting all Pods without changing the controller's bad image specification. The evidence required here is: The controller creates Pods on the previously working image and no longer restarts the faulty version.

The intended resource creates cleanly in the target cluster without conflicting immutable server-generated metadata. That observed result would support another specific correction: Remove server-managed identity and concurrency fields from portable declarative manifests before applying to another cluster. The other scenario concerns: A user copies a resource manifest between clusters while retaining server-generated uid and resourceVersion metadata. The current Timberline Foods problem is: An operator rolls back a broken release by deleting all Pods without changing the controller's bad image specification. The evidence required here is: The controller creates Pods on the previously working image and no longer restarts the faulty version.

A project user manages allowed resources in its namespace but cannot list restricted resources elsewhere. That observed result would support another specific correction: Use a least-privilege Role and RoleBinding scoped to the project namespace for ordinary namespaced operations. The other scenario concerns: A team gives every project full cluster administrator access because the namespace-specific Role syntax is unfamiliar. The current Timberline Foods problem is: An operator rolls back a broken release by deleting all Pods without changing the controller's bad image specification. The evidence required here is: The controller creates Pods on the previously working image and no longer restarts the faulty version.

Describe a ResourceQuota as enforcement of namespace consumption limits, not a node capacity provisioning request. That choice addresses a different foundational issue: A team assumes Kubernetes automatically provisions external cloud infrastructure for every ResourceQuota it creates. The issue at Juniper Media is instead: A training exercise suggests SSHing into nodes to modify running Pods directly instead of managing declarative resources. It does not establish the required correction; the decisive verification would be: A clean reconciliation reproduces the intended workload from versioned manifests without manual changes to nodes.

Use a declarative apply workflow for managed desired-state updates and confirm the resulting resource state. That choice addresses a different foundational issue: A new administrator applies a manifest using a command that only creates new objects and fails when the named resource already exists. The issue at Juniper Media is instead: A training exercise suggests SSHing into nodes to modify running Pods directly instead of managing declarative resources. It does not establish the required correction; the decisive verification would be: A clean reconciliation reproduces the intended workload from versioned manifests without manual changes to nodes.

Use the appropriate rollout status inspection and verify ready Pods before marking the release successful. That choice addresses a different foundational issue: A team changes a Deployment manifest but never checks whether rollout completed or became stuck. The issue at Juniper Media is instead: A training exercise suggests SSHing into nodes to modify running Pods directly instead of managing declarative resources. It does not establish the required correction; the decisive verification would be: A clean reconciliation reproduces the intended workload from versioned manifests without manual changes to nodes.

The active context, API endpoint, and changed resource match the intended environment in the audit trail. That observed result would support another specific correction: Inspect the active kubeconfig context and explicitly select the authorized cluster before applying changes. The other scenario concerns: An operator runs kubectl without selecting the intended cluster context and modifies a development workload by mistake. The current Juniper Media problem is: A training exercise suggests SSHing into nodes to modify running Pods directly instead of managing declarative resources. The evidence required here is: A clean reconciliation reproduces the intended workload from versioned manifests without manual changes to nodes.

The report distinguishes installed kubectl version from API server version and documents compatibility correctly. That observed result would support another specific correction: Check client and server version fields independently and follow supported version-skew guidance for client tooling. The other scenario concerns: An infrastructure group believes a kubectl client version mismatch automatically implies the cluster was upgraded. The current Juniper Media problem is: A training exercise suggests SSHing into nodes to modify running Pods directly instead of managing declarative resources. The evidence required here is: A clean reconciliation reproduces the intended workload from versioned manifests without manual changes to nodes.

Excess namespace resource creation is rejected by quota even though the cluster's physical nodes remain unchanged. That observed result would support another specific correction: Describe a ResourceQuota as enforcement of namespace consumption limits, not a node capacity provisioning request. The other scenario concerns: A team assumes Kubernetes automatically provisions external cloud infrastructure for every ResourceQuota it creates. The current Juniper Media problem is: A training exercise suggests SSHing into nodes to modify running Pods directly instead of managing declarative resources. The evidence required here is: A clean reconciliation reproduces the intended workload from versioned manifests without manual changes to nodes.

Use the Deployment's supported rollout undo or an explicitly reverted desired image version. That choice addresses a different foundational issue: An operator rolls back a broken release by deleting all Pods without changing the controller's bad image specification. The issue at Meadowbrook Finance is instead: A training exercise suggests SSHing into nodes to modify running Pods directly instead of managing declarative resources. It does not establish the required correction; the decisive verification would be: A clean reconciliation reproduces the intended workload from versioned manifests without manual changes to nodes.

Inspect the active kubeconfig context and explicitly select the authorized cluster before applying changes. That choice addresses a different foundational issue: An operator runs kubectl without selecting the intended cluster context and modifies a development workload by mistake. The issue at Meadowbrook Finance is instead: A training exercise suggests SSHing into nodes to modify running Pods directly instead of managing declarative resources. It does not establish the required correction; the decisive verification would be: A clean reconciliation reproduces the intended workload from versioned manifests without manual changes to nodes.

Use get for inventory and describe for resource details, events, and selected conditions when inspecting a resource. That choice addresses a different foundational issue: An engineer cannot distinguish kubectl get from kubectl describe and reads only a short status column during a fault. The issue at Meadowbrook Finance is instead: A training exercise suggests SSHing into nodes to modify running Pods directly instead of managing declarative resources. It does not establish the required correction; the decisive verification would be: A clean reconciliation reproduces the intended workload from versioned manifests without manual changes to nodes.

Authorized workload access still works while unauthorized ConfigMap reads no longer expose credentials. That observed result would support another specific correction: Use a Secret with appropriate access controls and remove the disclosed value from non-secret configuration and logs. The other scenario concerns: A workload team puts a secret into a public ConfigMap during troubleshooting and assumes the namespace hides it from all users. The current Meadowbrook Finance problem is: A training exercise suggests SSHing into nodes to modify running Pods directly instead of managing declarative resources. The evidence required here is: A clean reconciliation reproduces the intended workload from versioned manifests without manual changes to nodes.

The recorded output contains structured metadata and specification fields that can be reviewed without screenshots. That observed result would support another specific correction: Request the resource with kubectl get using a machine-readable YAML or JSON output format. The other scenario concerns: A platform analyst wants the exact YAML representation of an existing resource but repeatedly screenshots terminal output. The current Meadowbrook Finance problem is: A training exercise suggests SSHing into nodes to modify running Pods directly instead of managing declarative resources. The evidence required here is: A clean reconciliation reproduces the intended workload from versioned manifests without manual changes to nodes.

An authorization check grants the required API action to the intended identity and denies it to an unrelated account. That observed result would support another specific correction: Bind the appropriate Role or ClusterRole to the intended subject using a namespaced or cluster-scoped binding. The other scenario concerns: An administrator expects RBAC rules to automatically grant permissions merely by creating a Role with no RoleBinding. The current Meadowbrook Finance problem is: A training exercise suggests SSHing into nodes to modify running Pods directly instead of managing declarative resources. The evidence required here is: A clean reconciliation reproduces the intended workload from versioned manifests without manual changes to nodes.

Inspect the active kubeconfig context and explicitly select the authorized cluster before applying changes. That choice addresses a different foundational issue: An operator runs kubectl without selecting the intended cluster context and modifies a development workload by mistake. The issue at Eastshore Energy is instead: A resource list shows both cluster-scoped and namespace-scoped items and an operator attempts to fetch a Node in a namespace. It does not establish the required correction; the decisive verification would be: The Node listing works without a namespace filter while namespaced application resources remain scoped correctly.

Use get for inventory and describe for resource details, events, and selected conditions when inspecting a resource. That choice addresses a different foundational issue: An engineer cannot distinguish kubectl get from kubectl describe and reads only a short status column during a fault. The issue at Eastshore Energy is instead: A resource list shows both cluster-scoped and namespace-scoped items and an operator attempts to fetch a Node in a namespace. It does not establish the required correction; the decisive verification would be: The Node listing works without a namespace filter while namespaced application resources remain scoped correctly.

Use the appropriate rollout status inspection and verify ready Pods before marking the release successful. That choice addresses a different foundational issue: A team changes a Deployment manifest but never checks whether rollout completed or became stuck. The issue at Eastshore Energy is instead: A resource list shows both cluster-scoped and namespace-scoped items and an operator attempts to fetch a Node in a namespace. It does not establish the required correction; the decisive verification would be: The Node listing works without a namespace filter while namespaced application resources remain scoped correctly.

A project user manages allowed resources in its namespace but cannot list restricted resources elsewhere. That observed result would support another specific correction: Use a least-privilege Role and RoleBinding scoped to the project namespace for ordinary namespaced operations. The other scenario concerns: A team gives every project full cluster administrator access because the namespace-specific Role syntax is unfamiliar. The current Eastshore Energy problem is: A resource list shows both cluster-scoped and namespace-scoped items and an operator attempts to fetch a Node in a namespace. The evidence required here is: The Node listing works without a namespace filter while namespaced application resources remain scoped correctly.

Excess namespace resource creation is rejected by quota even though the cluster's physical nodes remain unchanged. That observed result would support another specific correction: Describe a ResourceQuota as enforcement of namespace consumption limits, not a node capacity provisioning request. The other scenario concerns: A team assumes Kubernetes automatically provisions external cloud infrastructure for every ResourceQuota it creates. The current Eastshore Energy problem is: A resource list shows both cluster-scoped and namespace-scoped items and an operator attempts to fetch a Node in a namespace. The evidence required here is: The Node listing works without a namespace filter while namespaced application resources remain scoped correctly.

Authorized workload access still works while unauthorized ConfigMap reads no longer expose credentials. That observed result would support another specific correction: Use a Secret with appropriate access controls and remove the disclosed value from non-secret configuration and logs. The other scenario concerns: A workload team puts a secret into a public ConfigMap during troubleshooting and assumes the namespace hides it from all users. The current Eastshore Energy problem is: A resource list shows both cluster-scoped and namespace-scoped items and an operator attempts to fetch a Node in a namespace. The evidence required here is: The Node listing works without a namespace filter while namespaced application resources remain scoped correctly.

Remove server-managed identity and concurrency fields from portable declarative manifests before applying to another cluster. That choice addresses a different foundational issue: A user copies a resource manifest between clusters while retaining server-generated uid and resourceVersion metadata. The issue at Alpine Healthcare is instead: A resource list shows both cluster-scoped and namespace-scoped items and an operator attempts to fetch a Node in a namespace. It does not establish the required correction; the decisive verification would be: The Node listing works without a namespace filter while namespaced application resources remain scoped correctly.

Use get for inventory and describe for resource details, events, and selected conditions when inspecting a resource. That choice addresses a different foundational issue: An engineer cannot distinguish kubectl get from kubectl describe and reads only a short status column during a fault. The issue at Alpine Healthcare is instead: A resource list shows both cluster-scoped and namespace-scoped items and an operator attempts to fetch a Node in a namespace. It does not establish the required correction; the decisive verification would be: The Node listing works without a namespace filter while namespaced application resources remain scoped correctly.

Use a Secret with appropriate access controls and remove the disclosed value from non-secret configuration and logs. That choice addresses a different foundational issue: A workload team puts a secret into a public ConfigMap during troubleshooting and assumes the namespace hides it from all users. The issue at Alpine Healthcare is instead: A resource list shows both cluster-scoped and namespace-scoped items and an operator attempts to fetch a Node in a namespace. It does not establish the required correction; the decisive verification would be: The Node listing works without a namespace filter while namespaced application resources remain scoped correctly.

The application's Pod appears under the expected namespace while the original default-namespace query remains empty. That observed result would support another specific correction: Specify the correct namespace or inspect all permitted namespaces when looking for a resource. The other scenario concerns: A user lists Pods without a namespace and concludes an application is missing although it runs in another namespace. The current Alpine Healthcare problem is: A resource list shows both cluster-scoped and namespace-scoped items and an operator attempts to fetch a Node in a namespace. The evidence required here is: The Node listing works without a namespace filter while namespaced application resources remain scoped correctly.

Authorized workload access still works while unauthorized ConfigMap reads no longer expose credentials. That observed result would support another specific correction: Use a Secret with appropriate access controls and remove the disclosed value from non-secret configuration and logs. The other scenario concerns: A workload team puts a secret into a public ConfigMap during troubleshooting and assumes the namespace hides it from all users. The current Alpine Healthcare problem is: A resource list shows both cluster-scoped and namespace-scoped items and an operator attempts to fetch a Node in a namespace. The evidence required here is: The Node listing works without a namespace filter while namespaced application resources remain scoped correctly.

A controlled individual-object deletion leaves unrelated resources untouched and the expected namespace persists. That observed result would support another specific correction: Differentiate deleting an individual object from deleting its parent namespace, and inspect namespace contents before cleanup. The other scenario concerns: A namespace cleanup removes one configuration object but leaves the namespace and its workload resources intact unexpectedly. The current Alpine Healthcare problem is: A resource list shows both cluster-scoped and namespace-scoped items and an operator attempts to fetch a Node in a namespace. The evidence required here is: The Node listing works without a namespace filter while namespaced application resources remain scoped correctly.

Use get for inventory and describe for resource details, events, and selected conditions when inspecting a resource. That choice addresses a different foundational issue: An engineer cannot distinguish kubectl get from kubectl describe and reads only a short status column during a fault. The issue at Brookside Manufacturing is instead: A team changes a resource by hand then loses the change on the next GitOps sync because it is absent from version control. It does not establish the required correction; the decisive verification would be: The subsequent GitOps sync retains the intended setting and reports no recurring configuration drift.

Use a server-side dry run or diff where supported and review proposed changes before applying the manifest. That choice addresses a different foundational issue: An operator runs a potentially destructive manifest update without seeing which objects will change. The issue at Brookside Manufacturing is instead: A team changes a resource by hand then loses the change on the next GitOps sync because it is absent from version control. It does not establish the required correction; the decisive verification would be: The subsequent GitOps sync retains the intended setting and reports no recurring configuration drift.

Inspect the Deployment's spec replicas and status fields separately from Service endpoint readiness. That choice addresses a different foundational issue: A cluster maintainer increases replicas but mistakes a Service's endpoint count for authoritative desired Deployment replicas. The issue at Brookside Manufacturing is instead: A team changes a resource by hand then loses the change on the next GitOps sync because it is absent from version control. It does not establish the required correction; the decisive verification would be: The subsequent GitOps sync retains the intended setting and reports no recurring configuration drift.

The intended resource creates cleanly in the target cluster without conflicting immutable server-generated metadata. That observed result would support another specific correction: Remove server-managed identity and concurrency fields from portable declarative manifests before applying to another cluster. The other scenario concerns: A user copies a resource manifest between clusters while retaining server-generated uid and resourceVersion metadata. The current Brookside Manufacturing problem is: A team changes a resource by hand then loses the change on the next GitOps sync because it is absent from version control. The evidence required here is: The subsequent GitOps sync retains the intended setting and reports no recurring configuration drift.

Unspecified requests obtain approved defaults and over-limit Pods or allocations are rejected predictably. That observed result would support another specific correction: Configure suitable LimitRanges and quotas to guide per-container resource defaults and total namespace budgets. The other scenario concerns: A project repeatedly allocates large default container requests because no namespace policy sets sensible boundaries. The current Brookside Manufacturing problem is: A team changes a resource by hand then loses the change on the next GitOps sync because it is absent from version control. The evidence required here is: The subsequent GitOps sync retains the intended setting and reports no recurring configuration drift.

The developer can read the intended namespace resources and receives a denial in a neighboring namespace. That observed result would support another specific correction: Bind an appropriate Role or ClusterRole through a RoleBinding in the target namespace instead of a global binding. The other scenario concerns: An operator grants a cluster-wide ClusterRoleBinding when a developer needs read access in just one namespace. The current Brookside Manufacturing problem is: A team changes a resource by hand then loses the change on the next GitOps sync because it is absent from version control. The evidence required here is: The subsequent GitOps sync retains the intended setting and reports no recurring configuration drift.