AI-200 Exam Dumps and Practice Questions

The solution does not meet the goal. Storing the API key as an App Service application setting through the Azure Portal violates the requirements in several ways:

Problems:

• App Service application settings are stored in plaintext or as encrypted configuration values, not in Azure Key Vault, which is the specified requirement.
• Application settings can be viewed in the portal and may appear in deployment templates or Git history if accidentally committed.
• The requirement explicitly states the API key must not be stored in application configuration directly.
• This approach does not leverage Key Vault's secure storage and secret rotation capabilities.

Correct approach: Use Azure Key Vault references in App Service configuration (format: @Microsoft.KeyVault(SecretUri=...)) combined with managed identity, or use Key Vault SDK in the application code to retrieve the secret at runtime.

Enable semantic ranking in Azure AI Search. Semantic ranking is a query-time secondary re-ranking stage that evaluates the initial candidate results using Microsoft language-understanding models and promotes documents that are more semantically aligned with the user's actual query intent. It can operate over BM25 results, hybrid results, and the textual content associated with vector-search results.

This directly addresses the stated problem. Vector similarity retrieval often has strong recall but can return passages that are broadly related rather than specifically useful for answering the question. Microsoft's RAG guidance explains that reranking improves precision by taking the retrieved candidate set and reordering it so that the most query-relevant chunks are placed first. This reduces irrelevant context passed to the generation model and improves grounding quality without retraining either the embedding model or the LLM.

Increasing embedding dimensionality alone does not guarantee better relevance. Disabling vector search sacrifices semantic retrieval capability, while reducing the index size arbitrarily removes potentially useful content rather than improving ranking quality.

Therefore, semantic re-ranking is the correct precision-improvement mechanism.

Study Guide references: Azure AI Search semantic ranker; RAG information retrieval; hybrid/vector search; secondary ranking; relevance optimization.

Detailed A KEDA custom scale rule must identify what scaler to use and which resource within that scaler should be observed. For Azure Storage queues, the trigger type selects the Storage Queue scaler and the queue name identifies the specific queue whose message count drives scaling. Polling interval and maximum replicas are general scaling controls, but they do not define the required queue-specific trigger. Therefore trigger type and queue name are the essential rule values tested here.

Study Guide Alignment: Containerized Azure workloads: registry builds, App Service containers, Container Apps revision/scaling behavior, and AKS deployment choices.

Official Microsoft Learn Reference: AI-200 Study Guide | Set scaling rules in Azure Container Apps

Detailed If the priority is to minimize stale results, source-driven invalidation is the direct control: when the underlying record changes, delete the corresponding cache entry so the next request repopulates it from current data. Sliding expiration can keep a frequently accessed stale item alive, disabling expiration is worse, and an allkeys-LRU policy evicts according to memory pressure rather than data freshness. Invalidation therefore best addresses correctness without requiring constant cache-wide maintenance.

Study Guide Alignment: AI data-management workloads: Cosmos DB, PostgreSQL, caching, vector storage, vector retrieval, consistency, and connection optimization.

Official Microsoft Learn Reference: AI-200 Study Guide | Azure Managed Redis documentation

The first troubleshooting action should be to inspect the pod events and container logs. Microsoft's AKS troubleshooting guidance identifies Kubernetes events and application logs as primary diagnostic sources when a pod repeatedly restarts or enters states such as CrashLoopBackOff. kubectl describe pod exposes events including failed liveness or readiness probes, restart reasons, scheduling problems, and container termination information. kubectl logs provides application-level output, while kubectl logs --previous is particularly useful when the container has already restarted because it retrieves logs from the previous terminated container instance.

The scenario already states that node CPU and memory are normal, making node resource pressure less likely. Probe failures can instead result from incorrect probe paths, ports, timeouts, dependency checks, slow application initialization, or application errors. Microsoft specifically recommends reviewing health probe configuration and application behavior before modifying infrastructure.

Scaling replicas does not identify the root cause. Decreasing initialDelaySeconds can actually make a slow-starting application fail sooner. Draining or rebooting the node is unjustified without evidence of a node-level fault.

Study Guide references: AKS troubleshooting Kubernetes events, container logs, liveness/readiness probes, kubectl describe, kubectl logs --previous.