Microsoft DP-700 Practice Test 33
Question 320
A Delta notebook fails reading mixed incompatible Parquet types from source files. What should be repaired?
- Normalize source schema or apply an explicit supported cast before the Delta read/write contract.
- Use a compatible checkpoint migration or a new checkpoint after validating replay and duplication risks.
- Replace driver-side collection with distributed DataFrame operations.
- Inspect skewed partition sizes and adjust partitioning or skew-handling strategy.
Explanation
Correct Answer
A:Physical schema inconsistencies can break inference and Delta compatibility; more executors cannot reconcile incompatible types. Inspect representative source schemas and rerun ingestion.
Incorrect Answers
B:Stateful checkpoint formats can be incompatible with structural query changes; blindly reusing the old location may fail.
C:Collect materializes rows in one driver process; keeping transformations distributed addresses the memory concentration.
D:A few oversized tasks can exhaust executors despite sufficient average cluster memory.
Question 321
An Eventhouse table shows zero newly ingested rows despite upstream event arrivals. Which diagnostic should be checked first?
- Inspect ingestion results and Eventhouse monitoring logs for errors on the target database.
- Replace the contains predicate with the KQL has operator to use the indexed complete-term lookup.
- Inspect the current table schema and adapt the query's column reference.
- Validate the shortcut target and connection permissions before rewriting the query.
Explanation
Correct Answer
A:Ingestion failures can prevent events becoming queryable even when the source connector is producing messages. Correlate source event time with ingestion result records.
Incorrect Answers
B:Reducing scanned event history can improve query efficiency without changing ingestion architecture.
C:KQL name resolution requires existing columns; an outdated field name is a query defect, not an ingestion-rate issue.
D:External tables depend on reachable source paths and authorization; the KQL syntax may be valid while storage access fails.
Question 322
A KQL query fails because it references a field removed from an event schema. What should the engineer update?
- Inspect ingestion results and Eventhouse monitoring logs for errors on the target database.
- Review the materialized view source definition and table dependencies.
- Normalize ingestion timestamp interpretation and query time boundaries consistently.
- Inspect the current table schema and adapt the query's column reference.
Explanation
Correct Answer
D:KQL name resolution requires existing columns; an outdated field name is a query defect, not an ingestion-rate issue. Run a small projection against the current schema.
Incorrect Answers
A:Ingestion failures can prevent events becoming queryable even when the source connector is producing messages.
B:A view cannot maintain output if its referenced source object no longer resolves to the intended table.
C:Timezone mismatch can make successfully ingested events appear absent from an expected window.
Question 323
An Eventhouse dashboard unexpectedly shows nulls after a source sends numbers as strings. Which boundary should be corrected?
- Repair the ingestion mapping or explicit KQL type conversion for the changed field.
- Inspect ingestion results and Eventhouse monitoring logs for errors on the target database.
- Review the materialized view source definition and table dependencies.
- Inspect the reported ingestion-error categories and source mapping details.
Explanation
Correct Answer
A:Type conversion errors can create nulls or rejected data; merely expanding the query time window won't correct types. Test valid and invalid sample values.
Incorrect Answers
B:Ingestion failures can prevent events becoming queryable even when the source connector is producing messages.
C:A view cannot maintain output if its referenced source object no longer resolves to the intended table.
D:Healthy storage doesn't prove event schema compatibility or connector success; targeted ingestion diagnostics are needed.
Question 324
A materialized view stopped updating after its source table was renamed. What should be inspected?
- Repair the ingestion mapping or explicit KQL type conversion for the changed field.
- Replace the contains predicate with the KQL has operator to use the indexed complete-term lookup.
- Validate the shortcut target and connection permissions before rewriting the query.
- Review the materialized view source definition and table dependencies.
Explanation
Correct Answer
D:A view cannot maintain output if its referenced source object no longer resolves to the intended table. Check the view status and source existence.
Incorrect Answers
A:Type conversion errors can create nulls or rejected data; merely expanding the query time window won't correct types.
B:Reducing scanned event history can improve query efficiency without changing ingestion architecture.
C:External tables depend on reachable source paths and authorization; the KQL syntax may be valid while storage access fails.
Question 325
An Eventhouse's system overview reports increasing ingestion errors but storage is healthy. What should the operator prioritize?
- Repair the ingestion mapping or explicit KQL type conversion for the changed field.
- Validate the shortcut target and connection permissions before rewriting the query.
- Normalize ingestion timestamp interpretation and query time boundaries consistently.
- Inspect the reported ingestion-error categories and source mapping details.
Explanation
Correct Answer
D:Healthy storage doesn't prove event schema compatibility or connector success; targeted ingestion diagnostics are needed. Compare sample failed events with mapping rules.
Incorrect Answers
A:Type conversion errors can create nulls or rejected data; merely expanding the query time window won't correct types.
B:External tables depend on reachable source paths and authorization; the KQL syntax may be valid while storage access fails.
C:Timezone mismatch can make successfully ingested events appear absent from an expected window.
Question 326
A KQL table is receiving events, but queries filter out all records because timestamps were parsed in the wrong timezone. What should be corrected?
- Review the materialized view source definition and table dependencies.
- Inspect the reported ingestion-error categories and source mapping details.
- Normalize ingestion timestamp interpretation and query time boundaries consistently.
- Inspect the current table schema and adapt the query's column reference.
Explanation
Correct Answer
C:Timezone mismatch can make successfully ingested events appear absent from an expected window. Query a broad absolute time range and compare raw timestamps.
Incorrect Answers
A:A view cannot maintain output if its referenced source object no longer resolves to the intended table.
B:Healthy storage doesn't prove event schema compatibility or connector success; targeted ingestion diagnostics are needed.
D:KQL name resolution requires existing columns; an outdated field name is a query defect, not an ingestion-rate issue.
Question 327
A newly configured external_table shortcut in a KQL database fails to resolve the referenced storage. What should be checked?
- Normalize ingestion timestamp interpretation and query time boundaries consistently.
- Inspect the current table schema and adapt the query's column reference.
- Validate the shortcut target and connection permissions before rewriting the query.
- Repair the ingestion mapping or explicit KQL type conversion for the changed field.
Explanation
Correct Answer
C:External tables depend on reachable source paths and authorization; the KQL syntax may be valid while storage access fails. Read a known source file with the effective identity.
Incorrect Answers
A:Timezone mismatch can make successfully ingested events appear absent from an expected window.
B:KQL name resolution requires existing columns; an outdated field name is a query defect, not an ingestion-rate issue.
D:Type conversion errors can create nulls or rejected data; merely expanding the query time window won't correct types.
Question 328
A Fabric Eventhouse query searches an entire message token FAILED by using KQL contains, and its text scan is slow. Only complete-token matches are needed. Which operator improves efficiency?
- Use startswith on message to find FAILED wherever it occurs in the message.
- Replace the contains predicate with the KQL has operator to use the indexed complete-term lookup.
- Use contains_cs instead of contains to keep searching for the same substring.
- Use matches regex to enumerate every occurrence of the term FAILED.
Explanation
Correct Answer
B:Reducing scanned event history can improve query efficiency without changing ingestion architecture. Compare scanned rows and duration with the original query.
Incorrect Answers
A:Startswith only finds a prefix and misses the required whole term appearing later inside a message.
C:Contains_cs changes case sensitivity but remains a substring match, which generally does not use the indexed full-term lookup needed by the scenario. The scenario, however, is that a Fabric Eventhouse query searches an entire message token FAILED by using KQL contains, and its text scan is slow.
D:Regex matching adds expression work and does not replace the efficient term index when a full-token literal match is sufficient. In the current incident, a Fabric Eventhouse query searches an entire message token FAILED by using KQL contains, and its text scan is slow.
Question 329
An Eventstream shows deserialization errors when a producer changes message format. What should be investigated?
- Inspect the filter predicate on the affected stream branch and its output counts.
- Correct field typing or transformation to handle invalid text-to-number conversions.
- Compare incoming payload encoding and schema against the configured source deserializer.
- Inspect destination-node Runtime logs and Data insights metrics.
Explanation
Correct Answer
C:Deserialization errors occur before downstream analytics when message bytes cannot be interpreted as configured. Replay one failing payload in a test stream.
Incorrect Answers
A:An overrestrictive filter can silently discard valid events even while upstream throughput remains healthy.
B:A conversion exception is about value/schema compatibility, not necessarily a missing network connection.
D:Node-specific logs distinguish output connector failure from source delivery, revealing which edge of the graph is broken.
Question 330
A streaming connector reports data-conversion errors for a numeric field. Which fix addresses the root cause?
- Compare incoming and outgoing throughput and processing errors for the affected nodes.
- Correct field typing or transformation to handle invalid text-to-number conversions.
- Renew the relevant source or destination connection authorization and resume the affected node.
- Inspect the filter predicate on the affected stream branch and its output counts.
Explanation
Correct Answer
B:A conversion exception is about value/schema compatibility, not necessarily a missing network connection. Test representative malformed numeric values.
Incorrect Answers
A:Backlog arises when processing or destination capacity lags arrival; node-level rates identify the limiting stage.
C:Authentication failures prevent connector operations; throughput tuning cannot authorize a revoked identity.
D:An overrestrictive filter can silently discard valid events even while upstream throughput remains healthy.
