Selected foundations; all 30 retained official leaves remain partial. Additional evidence is not a full-outline or calibrated simulation claim.. Selected foundation practice, not a complete exam simulation.
Try 10 original samples with answers below, or start a personal practice set.
Free access · No sign-up · No personal progress recorded on this page · Coverage & review notes
TRY A QUESTION. FOLLOW THE REASONING.
Sample questions with answers
Work out your answer first, then compare the explanation. This stable selection uses distinct task families across the available learning domains.
QUESTION 01Core data concepts
Data representations
A fictional research team is sorting five representations. Select TWO that the descriptions identify as semi-structured, rather than fixed-schema tabular data or unstructured media.
Select all answers that apply.
JSON device records whose optional property names vary from device to device
A CSV export with the same declared columns for every row
Tagged event records whose permitted payload fields vary by event type
Audio recordings stored as binary media files
A SQL table whose rows use the same declared typed columns
Answer: JSON device records whose optional property names vary from device to device; Tagged event records whose permitted payload fields vary by event type
Classify the representation described, rather than deciding from the product name alone. The varying optional properties and event-specific payload fields both provide organization without one fixed set of fields for every record.
The declared CSV columns and SQL columns define fixed tabular schemas in these descriptions. Audio recordings are the unstructured media example. JSON can be used with a fixed schema too; the stated field variation controls this case.
Check a changed condition: if the event contract instead required exactly the same declared fields on every record, variable payloads would no longer be the stated reason to call it semi-structured. The answer is tied to the descriptions, not to a permanent ranking of file extensions. Audio metadata can be tabular without making the audio itself a fixed-schema record.
Compare the other answer choices
A CSV export with the same declared columns for every row
The CSV description explicitly fixes the columns across rows; human-readable text does not by itself make that representation semi-structured.
Audio recordings stored as binary media files
The audio description supplies media content rather than field-bearing records with a flexible organization.
A SQL table whose rows use the same declared typed columns
The SQL table description uses the same declared columns for each row, which is the fixed-schema contrast.
A complete fictional table Loans(LoanId, Days, Status) has four rows: (A, 2, 'Open'); (B, 6, 'Closed'); (C, 5, 'Open'); (D, 8, 'Open'). There are no NULL values. How many rows satisfy SELECT LoanId FROM Loans WHERE Status = 'Open' AND Days >= 5?
3 rows
2 rows
1 row
4 rows
Answer: 2 rows
Check both predicates for each row: A fails Days >= 5; B fails Status = 'Open'; C passes both, including the equality boundary; D passes both.
The result contains C and D, so the count is 2. Ignoring Status counts B, C and D (3). Replacing >= with > keeps only D (1). Replacing AND with OR admits A, B, C and D (4).
Enumerate each supplied row against both predicates: A is Open but below five; B meets the numeric threshold but is Closed; C is Open at the inclusive boundary; D is Open above it. Only C and D survive. The boundary word >= matters independently of AND. Replacing AND with OR would admit rows satisfying only one condition and answer a different selection request.
Compare the other answer choices
3 rows
Ignoring the status predicate counts B, C and D: 3 rows. B is closed and must be excluded.
1 row
Changing Days >= 5 into Days > 5 wrongly drops C at the equality boundary and leaves D: 1 row.
4 rows
Using OR instead of AND includes all four rows: A passes status, B passes duration, and C/D pass both.
Two fictional applications must retain their current storage interfaces without an application rewrite. A requires an SMB shared-file endpoint. B already issues Azure Blob REST API requests to store and retrieve media objects. Select TWO proposals that match the respective stated interfaces.
Select all answers that apply.
Use Azure Files over SMB for application A
Use Azure Table Storage as application A’s existing SMB share
Use Azure Blob Storage for application B’s existing Blob REST requests
Use a Blob NFS interface as application A’s unchanged SMB endpoint
Use an Azure Files endpoint for application B’s unchanged Blob REST requests
Answer: Use Azure Files over SMB for application A; Use Azure Blob Storage for application B’s existing Blob REST requests
Keep the two interfaces separate: A needs SMB file sharing; B needs the Blob REST interface for object operations. Azure Files addresses the former, while Blob Storage addresses the latter.
A file-like capability is not automatically the same wire protocol. Blob Storage supports NFS in applicable configurations, but that does not turn NFS into the required SMB endpoint. Likewise, an Azure Files REST endpoint is not the unchanged Blob REST interface.
Test each application's actual endpoint rather than choosing one service for the broad word files. SMB is the first unchanged interface; Blob REST is the second. A supported alternative protocol does not automatically accept the requests already sent by either client. If an application rewrite were allowed, the candidates might change, but the present comparison intentionally fixes the client contracts rather than comparing all storage uses.
Compare the other answer choices
Use Azure Table Storage as application A’s existing SMB share
Table Storage exposes a key/attribute NoSQL model, not the existing SMB file-share interface required by A.
Use a Blob NFS interface as application A’s unchanged SMB endpoint
NFS and SMB are different protocols. A explicitly requires SMB without rewriting the application.
Use an Azure Files endpoint for application B’s unchanged Blob REST requests
Files can expose REST access, but B specifically retains Blob REST requests; generic REST support does not make the service APIs interchangeable.
A fictional pipeline extracts source data, loads the target analytical store, and then transforms the loaded data there. Which ordered expansion names that pattern?
ELT:extract, load, transform
ETL:extract, transform, load
LTE:load, transform, extract
TLE:transform, load, extract
Answer: ELT:extract, load, transform
Preserve the order in the description. Extraction occurs first, target loading second, and target-side transformation third.
That sequence is ELT. ETL changes the loading/transformation order; the other proposed permutations do not preserve the specified process.
Name the physical point where the transform is executed. If source data is loaded into the analytical target first and then transformed there, L precedes T even if extraction also performs transport preparation. ETL would describe a different ordering in which the relevant transformation precedes target loading. An acronym is useful only after the described operations and target boundary have been identified.
Compare the other answer choices
ETL:extract, transform, load
ETL performs transformation before target loading, not afterward.
LTE:load, transform, extract
This moves extraction to the end, contrary to the stated process.
TLE:transform, load, extract
This starts with transformation rather than extracting the stated source input.
QUESTION 05Relational and non-relational decisions
Service contracts and transaction boundaries
In a fictional Cosmos DB for NoSQL single-region-write workflow, two clients read item version E1. Client A conditionally replaces the item and it becomes E2. Client B must avoid silently overwriting A's change with its old calculation. Which next request-and-response interpretation fits optimistic concurrency?
Keep the local E1 value but compare it only with the E1 value in B’s original cache
Send B’s replacement with If-Match E1; treat a precondition failure as a reason to reread and reconcile
Read E2 and attach it to B’s unchanged old calculation without reviewing the changed fields
Send B’s replacement without a version condition and use the write success as evidence that no concurrent edit occurred
Answer: Send B’s replacement with If-Match E1; treat a precondition failure as a reason to reread and reconcile
The comparison must involve the current server item. B's If-Match E1 condition no longer matches E2, so the conditional write is rejected with a precondition failure rather than silently replacing the newer version. The client should fetch the current item and decide how to reconcile the business change.
Merely obtaining the new tag and applying an old calculation defeats the intended lost-update protection. Likewise, an unconditional successful write only says that a write was accepted; it does not show that B calculated from the latest state. The stem limits this case to the described single-region workflow rather than generalizing every multi-region conflict policy.
Compare the other answer choices
Read E2 and attach it to B’s unchanged old calculation without reviewing the changed fields
Refreshing a token without recomputing or reconciling the old business update can still discard A’s changed fields.
Send B’s replacement without a version condition and use the write success as evidence that no concurrent edit occurred
Omitting the condition removes the stated version check; success does not certify that no intervening write happened.
Keep the local E1 value but compare it only with the E1 value in B’s original cache
Two cached copies of the same old value do not test the server’s current version.
Reference basis: Microsoft Learn: rd-dp-cosmos. These are original learning questions, not questions reproduced from the references.
QUESTION 06Analytics workloads
Analytical grain, timing and platform evidence
A fictional service needs both a running arrival count during the day and a reproducible daily summary from the complete retained day’s inputs. Select TWO design statements that address these separate output contracts without claiming a service guarantee.
Select all answers that apply.
Define the daily input boundary and retained input set so the daily calculation can be rerun for that set
Compute and publish an ongoing aggregate from arriving events for the running view
Use the existence of an ingestion buffer as evidence that the running aggregate has already been computed
Treat the first running result as the final daily result without a daily completeness rule
Use the same code name for both outputs and infer that their freshness and completeness contracts match
Answer: Compute and publish an ongoing aggregate from arriving events for the running view; Define the daily input boundary and retained input set so the daily calculation can be rerun for that set
The running view needs computation over incoming events, not just transport or storage. The reproducible daily view additionally needs a named bounded input set so another run can use the same data. These outputs can share logic or infrastructure while retaining different freshness and completeness contracts.
A running count observed midday is not necessarily the final day's count. A retained archive supports reconstruction only when the included input set and computation are defined. The answer does not promise zero latency, absence of duplicates or exactly-once delivery; those would need separate explicit design and evidence.
Compare the other answer choices
Use the existence of an ingestion buffer as evidence that the running aggregate has already been computed
Buffering arrivals does not by itself execute aggregation or publish a result.
Treat the first running result as the final daily result without a daily completeness rule
The first running view precedes the stated daily boundary and lacks evidence that the day’s inputs are complete.
Use the same code name for both outputs and infer that their freshness and completeness contracts match
Names do not establish input boundaries or output guarantees; the two tasks have explicitly different timing requirements.
A complete fictional CSV export is described by four lines: header ClientId, Units; then C1,3; then C2,7; then C3,0. Each data line is one client record. How many client records does this export contain?
4 records
10 records
3 records
2 records
Answer: 3 records
Separate the header from the data lines. C1, C2 and C3 are the three records; the zero Units value does not remove C3.
Counting the header gives4. Adding Units gives3+7+0=10 units rather than a record count. Counting the two column names gives2 columns.
Mark the header as schema information, then enumerate C1, C2 and C3 as the three records. C3's zero Units is still a supplied client record; omitting it changes the population from clients to clients with positive units. Counting four would instead mix one header line with three data records. This is an explicit simple export model, not a claim that every physical CSV line is a complete record.
Compare the other answer choices
4 records
This counts the header as though it were another client.
10 records
This sums Units,3+7+0=10, which is not the requested count.
2 records
This counts the two fields, not the three data lines.
Select TWO offerings described as managed SQL Server PaaS database platforms, where underlying OS/database maintenance is delegated as part of that platform rather than a guest VM being administered by the customer.
Select all answers that apply.
Azure SQL Database
SQL Server on a customer-administered Azure VM
SQL Server on the organization’s own physical server
SQL Database and Managed Instance provide managed SQL Server database platforms. Their common maintenance boundary does not mean their supported feature scopes are identical.
Running database software on a self-administered VM or physical host is a different operating model. Adding the word Azure to that host does not make it one of these two managed SQL Server PaaS offerings.
Keep the responsibility boundary explicit: delegating underlying maintenance does not delegate every data definition, access decision, query design or business recovery requirement. The two named PaaS offerings satisfy the question's managed-platform description, while the customer-administered guest VM does not. A service supporting the SQL Server engine can belong to a different operating model; matching the engine name alone does not settle who maintains its host.
Compare the other answer choices
SQL Server on a customer-administered Azure VM
The stated VM is customer-administered; it is not the managed database-platform boundary in the question.
SQL Server on the organization’s own physical server
A customer physical server does not supply the requested Azure PaaS platform.
PostgreSQL on a customer-administered Azure VM
The option explicitly specifies a customer-administered VM and a different engine.
A fictional design stores one JSON document per venue. Each document contains that venue’s current notes as a nested collection, and is identified by VenueId. Which request directly matches retrieving one whole venue record by its identifier, rather than combining multiple venue records?
Count notes across all venues grouped by year
Retrieve the current nested notes for a specified VenueId
Combine a separately maintained author profile with every venue that mentions it
Find pairs of venues that share authors across the complete collection
Answer: Retrieve the current nested notes for a specified VenueId
The stated grouping is one venue and its nested notes per identified document. Retrieving that venue’s current notes uses the one-record grouping directly.
The other requests concern collection-wide grouping or cross-record relationships. They may be valid needs, but this one-document retrieval fact does not by itself describe how to implement them.
Draw one document boundary around a specified venue and its nested notes. The requested read can follow that identifier to one grouped record. A collection-wide comparison may also be a legitimate business operation but would need a different access strategy. The example therefore identifies a direct fit for the stated grouping; it does not prove documents are always better than relations or that every cross-document question is impossible.
Compare the other answer choices
Count notes across all venues grouped by year
A collection-wide annual count combines information from multiple venue records.
Combine a separately maintained author profile with every venue that mentions it
That request relates separately maintained profiles to multiple venue records, not one whole venue record.
Find pairs of venues that share authors across the complete collection
Finding shared-author pairs compares multiple venues rather than retrieving one identified venue document.
A fictional device-monitoring task must update a running result as readings arrive throughout the day, rather than wait for a daily or weekly collection to finish. Which processing pattern directly matches that arrival requirement?
Process incoming readings as an ongoing event stream
Wait for the nightly file to close and run a daily batch
Build a weekly snapshot after all seven daily files arrive
Perform one historical backfill after the collection campaign ends
Answer: Process incoming readings as an ongoing event stream
The task describes ongoing arrivals and ongoing updates, so stream processing is the direct match. A bounded nightly or weekly batch waits for a collection boundary that the stated task rejects.
Streaming is a processing pattern, not a promise of zero latency or proof that every event is handled correctly. A production design would still need explicit timing, failure and data-quality requirements.
The arrival requirement is ongoing publication, so waiting for the stated nightly or weekly boundary changes the output contract. Streaming still needs real computation, error handling and a definition of lateness; choosing the pattern does not supply those guarantees. A design may also retain events for a later bounded daily reconciliation, but that additional batch output would complement rather than replace the running view requested here.
Compare the other answer choices
Wait for the nightly file to close and run a daily batch
A nightly batch waits for the day’s collection to close, contrary to the requested ongoing updates.
Build a weekly snapshot after all seven daily files arrive
The weekly snapshot delays processing until multiple completed collections exist, widening the mismatch.
Perform one historical backfill after the collection campaign ends
A historical backfill handles a completed past collection; it does not maintain the current running result while readings arrive.
Selected foundations; all 30 retained official leaves remain partial. Additional evidence is not a full-outline or calibrated simulation claim.
Learn data representations, workloads, Azure service boundaries and analytical reasoning with original cases and supplied small datasets.
53 single-choice and 11 select-TWO original learning tasks. These formats are not an official-weighted or calibrated full-length simulation.
10 public examples64 question variants in the studio42 task families
Review status: Independent educator review and learner calibration are pending. Use these selected foundation examples alongside the coverage map and current official references.
Start with the lesson linked to an unfamiliar topic, try a focused set in the practice studio, then return to the review cards. Your practice progress stays in this browser.
Catalog scope record: . This records the saved content mapping, not a guarantee that every external rule is current.
64 original formal tasks; related-concept reinforcement is included and is not 64 independent competencies.
14 lessons, 42 retrieval cards and 4 independent guides support selected foundations; every retained official leaf remains partial.
The 14-day route links real learning objects; the short worksheet reuses 8 tasks.
No official/recalled exam items, score prediction or provider/tenant deployment certification.
Professional review pending; representative learner trial not_run; release HOLD.
Check the exam provider's current information for registration, test rules and official preparation materials. The exam guide collects the relevant official references.
