CertifyDeckCargando contenido…
DP-700 · Sample questions
10 real questions from the CertifyDeck DP-700 bank, spread across the domains of the official Microsoft Certified: Fabric Data Engineer Associate skills outline, with the correct answer and a full explanation of why the other options are wrong. They are original questions written from the public syllabus, not exam dumps, and they are the same style you will find in the practice tests and mock exams.
Correct answer: B. Turn on shortcut caching in the OneLake settings of WS-Retail.
Shortcut caching, enabled in the OneLake tab of the workspace settings, stores files read through external shortcuts such as S3 in a workspace cache and serves later reads from it, which reduces cross-cloud egress. A scheduled Copy activity would copy the files, which the requirement rules out. OneLake diagnostics only logs data access, and OPTIMIZE compacts Delta files without avoiding repeated remote reads.
Correct answer: A. A default lakehouse rule for the notebook in the Test stage
Default lakehouse rules apply only to notebooks and set which lakehouse is connected as the notebook's default in the target stage, where the rule is defined. Data source and parameter rules are available for items such as semantic models, dataflows Gen1, paginated reports and mirrored databases, not for notebooks. Lakehouses aren't among the items you can create deployment rules for.
Correct answer: B. Edit or delete the DefaultReader role of LH_Flota.
Every lakehouse has a DefaultReader role whose members are all users with ReadAll, and it grants Read on all data; because roles combine as a union, Grp_BI keeps full access through it. Editing or deleting DefaultReader removes that access while item permissions stay the same. Deny roles don't exist, and adding more grants can't reduce access.
Correct answer: D. Domain assignment doesn't change access; use workspace roles and item permissions.
Assigning a workspace to a domain adds a domain attribute to its items for discovery and delegated governance, but it doesn't affect item visibility or accessibility; access depends on workspace roles and item permissions. Overriding certification settings only controls who can certify items in the domain. A default domain automates workspace assignment, and domain contributors are workspace admins allowed to assign their workspaces, so neither restricts who can open items.
Correct answer: C. The queryinsights.long_running_queries view in the warehouse.
Query insights is built into the warehouse and keeps historical query data; the queryinsights.long_running_queries view returns queries by execution time, which identifies the slowest transformations. queryinsights.frequently_run_queries focuses on how often queries run, not how long they take. Spark resource graphs and Dataflow Gen2 refresh history don't record T-SQL statements executed by warehouse stored procedures.
Correct answer: D. .show ingestion failures | where FailedOn > ago(1d)
.show ingestion failures returns failures of data ingestion management commands such as .ingest into, with a Details column for the root cause and FailureKind showing Permanent or Transient. .show streamingingestion failures only covers streaming ingestion, and queries over the Sales table only show rows that were ingested, not the files that failed.
Correct answer: C. Remove the data destination from the query, add it again and republish the dataflow.
DestinationQueryNotFound means the internal query that holds the navigation steps to the destination is missing, and the documented fix is to remove the data destination, add it again and publish. Reauthenticating addresses DestinationQueryChallenge, fixing the return type addresses DestinationQueryHasUnsupportedScript, and removing duplicates addresses MultipleDestinationQueries.
Correct answer: D. It deletes the data in the bucket's range and inserts the bucket again.
Incremental refresh in Dataflow Gen2 retrieves the whole bucket whose maximum value changed and uses a replace approach: it deletes the destination data for that bucket's time range and inserts the fresh data, leaving data outside the range untouched. It does not perform a key-based merge of individual rows, and it does not truncate and reload the full table.
Correct answer: D. CDC-based incremental copy with the Merge update method.
With CDC enabled on the source, a Copy job can use CDC-based incremental copy, and merging into the destination inserts, updates or deletes rows to match the source. Watermark-based incremental copy compares an incremental column with the last run, so it only tracks inserts and updates and never sees deleted rows, and a full copy moves every row on each run instead of only the changes.
Correct answer: D. Use writeStream.foreachBatch with a function that runs MERGE INTO customers.
Merge operations into Delta Lake from a streaming query are done with foreachBatch, which applies a batch function, such as MERGE INTO, to the output of every micro-batch. Update output mode looks tempting, but a Delta sink supports append and complete modes, not update. Complete mode only works with streaming aggregations and rewrites the result, and append mode with any trigger only adds rows, so existing customers would be duplicated instead of updated.
Answer each question before opening the explanation, and write down why you discarded the other options. On DP-700, most mistakes come from choosing an option that is technically possible but does not meet every requirement in the scenario.
These 10 questions are a taste of the format, not a diagnostic. To know which domains you fail, practice by domain for free and take a timed mock exam that mirrors the official weights and duration.
Start for free, no sign-up or card. Every question includes an explanation like the ones above, and the mock exam reproduces the official DP-700 format.