Question 510 A customer incident dashboard requires new notes searchable within minutes, yet the current search index refreshes only once nightly. What prevention measure is most targeted to this underlying cause? Monitor end-to-end ingestion-to-retrieval age with SLO alerts. Document source-update frequency and revisit index refresh settings when the business freshness requirement changes. Record index type and sync responsibilities in the data-index contract before launch. Revisit storage, vector count and endpoint utilization quarterly, with explicit headroom policy. ExplanationCorrect Answer A:Search freshness is bounded by synchronization lag; a nightly index cannot satisfy an intraday response SLO. Other Answers B:This control addresses a different exposure: A document corpus updates once each quarter, but an engineer configures very frequent index refreshes that dominate operating cost. It would not prevent this failure, because search freshness is bounded by synchronization lag; a nightly index cannot satisfy an intraday response SLO. C:That search freshness is bounded by synchronization lag; a nightly index cannot satisfy an intraday response SLO. D:Search freshness is bounded by synchronization lag; a nightly index cannot satisfy an intraday response SLO.
Question 511 A team selects an extremely large vector index with excessive replicas for a small low-traffic knowledge base and sees high fixed infrastructure charges. Which implementation decision best restores the intended behavior? Choose a sync cadence proportional to the actual source-change SLA rather than the minimum technically possible refresh interval. Right-size supported index/endpoint capacity based on actual embeddings, query concurrency and latency measurements. Use an appropriate supported Databricks Foundation Model API pay-per-token or provisioned throughput endpoint based on the measured traffic profile. Use a supported more frequent or continuous synchronization configuration consistent with the ingestion and search latency budget. ExplanationCorrect Answer B:The action targets the stated cause: Provisioned search infrastructure can carry fixed cost even when traffic is low; capacity should reflect observed demand. Check the repair by: load-test representative concurrency tiers and record P95 search latency plus cost before choosing capacity. Other Answers A:Provisioned search infrastructure can carry fixed cost even when traffic is low; capacity should reflect observed demand. C:Provisioned search infrastructure can carry fixed cost even when traffic is low; capacity should reflect observed demand. D:Provisioned search infrastructure can carry fixed cost even when traffic is low; capacity should reflect observed demand.
Question 512 A team selects an extremely large vector index with excessive replicas for a small low-traffic knowledge base and sees high fixed infrastructure charges. Which underlying technical principle explains why this occurs? Search freshness is bounded by synchronization lag; a nightly index cannot satisfy an intraday response SLO. Index synchronization incurs compute and update overhead; freshness has little value when source content rarely changes. Managed foundation-model endpoints remove many custom-hosting concerns but throughput, latency and cost still vary by serving mode. Provisioned search infrastructure can carry fixed cost even when traffic is low; capacity should reflect observed demand. ExplanationCorrect Answer D:Right-size supported index/endpoint capacity based on actual embeddings, query concurrency and latency measurements. Other Answers A:That mechanism describes another problem: A customer incident dashboard requires new notes searchable within minutes, yet the current search index refreshes only once nightly. It does not explain this case, where provisioned search infrastructure can carry fixed cost even when traffic is low; capacity should reflect observed demand. B:Provisioned search infrastructure can carry fixed cost even when traffic is low; capacity should reflect observed demand. C:Provisioned search infrastructure can carry fixed cost even when traffic is low; capacity should reflect observed demand.
Question 513 A conversational agent keeps raw OAuth tokens and complete customer records in its persistent memory table for convenience. What is the most targeted corrective step for this observed failure? Use approved server-side credentials and tool authentication with scoped network access, never model-visible plaintext secrets. Persist essential typed session state in an authorized durable store keyed by session and user identity, then resume safely. Store only approved non-sensitive conversation state and keep secrets in managed credential storage with short-lived tool access. Scope durable memory to authenticated user/tenant identity and use unguessable session identifiers with server-side ACL enforcement. ExplanationCorrect Answer C:The action targets the stated cause: Persisted memory expands the exposure and retention of sensitive data; tool authentication is separate from session context. Check the repair by: inspect stored state after a simulated conversation and confirm no token or protected record remains. Other Answers A:Persisted memory expands the exposure and retention of sensitive data; tool authentication is separate from session context. B:Persisted memory expands the exposure and retention of sensitive data; tool authentication is separate from session context. D:Persisted memory expands the exposure and retention of sensitive data; tool authentication is separate from session context.
Question 514 A conversational agent keeps raw OAuth tokens and complete customer records in its persistent memory table for convenience. What is the decisive mechanism behind the observed result? Process memory is ephemeral and fails across restarts or scaled replicas; durable state requires explicit storage and authorization. Persisted memory expands the exposure and retention of sensitive data; tool authentication is separate from session context. MCP tool descriptions and model prompts are not secret-management boundaries; credentials belong in the service runtime. A session identifier alone is not proof of authorization; memory reads must check identity before returning state. ExplanationCorrect Answer B:Store only approved non-sensitive conversation state and keep secrets in managed credential storage with short-lived tool access. Other Answers A:Persisted memory expands the exposure and retention of sensitive data; tool authentication is separate from session context. C:That mechanism describes another problem: An engineering team hosts a custom MCP server in a Databricks App but passes API keys through plain prompt text rather than through authenticated tool transport. It does not explain this case, where persisted memory expands the exposure and retention of sensitive data; tool authentication is separate from session context. D:Persisted memory expands the exposure and retention of sensitive data; tool authentication is separate from session context.
Question 515 A conversational agent keeps raw OAuth tokens and complete customer records in its persistent memory table for convenience. Which verification is strongest for this particular failure mode? Inspect sampled tool traces and outbound requests to verify no API secrets appear in model context. Restart an instance mid-workflow and verify the next request recovers only the permitted session phase and record ID. Attempt cross-user access with a stolen session ID and verify denial plus audit evidence. Inspect stored state after a simulated conversation and confirm no token or protected record remains. ExplanationCorrect Answer D:Persisted memory expands the exposure and retention of sensitive data; tool authentication is separate from session context. Other Answers A:Inspect stored state after a simulated conversation and confirm no token or protected record remains. B:Inspect stored state after a simulated conversation and confirm no token or protected record remains. C:Inspect stored state after a simulated conversation and confirm no token or protected record remains.
Question 516 A conversational agent keeps raw OAuth tokens and complete customer records in its persistent memory table for convenience. What change to engineering governance would reduce this risk over time? Apply field-level retention rules, encryption and memory redaction tests. Require secret scanning, short-lived credentials and egress controls for custom MCP services. Version session schemas and test recovery, expiration and ownership transitions. Enforce tenant-bound keys and periodic session revocation tests. ExplanationCorrect Answer A:Persisted memory expands the exposure and retention of sensitive data; tool authentication is separate from session context. Other Answers B:This control addresses a different exposure: An engineering team hosts a custom MCP server in a Databricks App but passes API keys through plain prompt text rather than through authenticated tool transport. It would not prevent this failure, because persisted memory expands the exposure and retention of sensitive data; tool authentication is separate from session context. C:That persisted memory expands the exposure and retention of sensitive data; tool authentication is separate from session context. D:Persisted memory expands the exposure and retention of sensitive data; tool authentication is separate from session context.
Question 517 A customer calls back after a service restart, but the agent's in-process dictionary lost the order_id and current workflow phase. Which engineering action most directly addresses this specific requirement? Scope durable memory to authenticated user/tenant identity and use unguessable session identifiers with server-side ACL enforcement. Add staging integration tests with the same scoped service-principal grants before production promotion. Persist essential typed session state in an authorized durable store keyed by session and user identity, then resume safely. Store only approved non-sensitive conversation state and keep secrets in managed credential storage with short-lived tool access. ExplanationCorrect Answer C:The action targets the stated cause: Process memory is ephemeral and fails across restarts or scaled replicas; durable state requires explicit storage and authorization. Check the repair by: restart an instance mid-workflow and verify the next request recovers only the permitted session phase and record ID. Other Answers A:Process memory is ephemeral and fails across restarts or scaled replicas; durable state requires explicit storage and authorization. B:Process memory is ephemeral and fails across restarts or scaled replicas; durable state requires explicit storage and authorization. D:Process memory is ephemeral and fails across restarts or scaled replicas; durable state requires explicit storage and authorization.
Question 518 A customer calls back after a service restart, but the agent's in-process dictionary lost the order_id and current workflow phase. What architectural constraint is most directly responsible? Local unit tests cannot establish real workspace permissions or endpoint accessibility under deployment identity. Persisted memory expands the exposure and retention of sensitive data; tool authentication is separate from session context. A session identifier alone is not proof of authorization; memory reads must check identity before returning state. Process memory is ephemeral and fails across restarts or scaled replicas; durable state requires explicit storage and authorization. ExplanationCorrect Answer D:Persist essential typed session state in an authorized durable store keyed by session and user identity, then resume safely. Other Answers A:That mechanism describes another problem: A Git workflow deploys an agent without running tests against the target workspace's security permissions, causing launch-time authorization failures. It does not explain this case, where process memory is ephemeral and fails across restarts or scaled replicas; durable state requires explicit storage and authorization. B:Process memory is ephemeral and fails across restarts or scaled replicas; durable state requires explicit storage and authorization. C:Process memory is ephemeral and fails across restarts or scaled replicas; durable state requires explicit storage and authorization.
Question 519 A customer calls back after a service restart, but the agent's in-process dictionary lost the order_id and current workflow phase. What validation evidence would most specifically confirm the resolution? Inspect stored state after a simulated conversation and confirm no token or protected record remains. Attempt cross-user access with a stolen session ID and verify denial plus audit evidence. Execute allowed and forbidden model/index access tests in staging and verify the expected audit identities. Restart an instance mid-workflow and verify the next request recovers only the permitted session phase and record ID. ExplanationCorrect Answer D:Process memory is ephemeral and fails across restarts or scaled replicas; durable state requires explicit storage and authorization. Other Answers A:Restart an instance mid-workflow and verify the next request recovers only the permitted session phase and record ID. B:Restart an instance mid-workflow and verify the next request recovers only the permitted session phase and record ID. C:Restart an instance mid-workflow and verify the next request recovers only the permitted session phase and record ID.
Question 520 A customer calls back after a service restart, but the agent's in-process dictionary lost the order_id and current workflow phase. What prevention measure is most targeted to this underlying cause? Version session schemas and test recovery, expiration and ownership transitions. Enforce tenant-bound keys and periodic session revocation tests. Make environment-specific least-privilege checks a required CI promotion gate. Apply field-level retention rules, encryption and memory redaction tests. ExplanationCorrect Answer A:Process memory is ephemeral and fails across restarts or scaled replicas; durable state requires explicit storage and authorization. Other Answers B:That process memory is ephemeral and fails across restarts or scaled replicas; durable state requires explicit storage and authorization. C:Process memory is ephemeral and fails across restarts or scaled replicas; durable state requires explicit storage and authorization. D:This control addresses a different exposure: A conversational agent keeps raw OAuth tokens and complete customer records in its persistent memory table for convenience. It would not prevent this failure, because process memory is ephemeral and fails across restarts or scaled replicas; durable state requires explicit storage and authorization.
