Question 122 of 820
Describe considerations for working with non-relational data on Azure →mediumMultiple ChoiceObjective-mapped
DP-900 Practice Question: Describe considerations for working with non-relational data on Azure
A global social media startup stores user profiles as JSON documents in Azure Cosmos DB. Their application frequently reads profiles by user ID and also runs queries to find users based on location or interests. The workload is read-heavy with high throughput requirements. The operations team notices that query performance degrades during peak hours. Which action would most effectively improve query performance?
⚠ Common exam trap
Watch out — candidates often confuse throughput (RU/s) with other scaling mechanisms like partitioning or API choice, but the core issue in a read-heavy, high-throughput scenario is insufficient provisioned capacity, not data organization or protocol differences.
Answer choices
Why each option matters
Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.
Correct answer & explanation
✓
Increase the provisioned throughput (RU/s)
Increasing the provisioned throughput (RU/s) directly allocates more processing capacity to the Cosmos DB container, allowing it to handle higher request volumes and reduce throttling during peak hours. Since the workload is read-heavy and query performance degrades under high throughput demands, raising RU/s is the most effective and immediate action to improve performance.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
Increase the number of containers
Why it's wrong here
Increasing the number of containers in a Cosmos DB database does not, by itself, add any Request Unit capacity. Containers created under a database share the database's provisioned throughput, and containers with dedicated throughput use their own separate allocation. Without increasing the total provisioned RU/s, splitting data across more containers simply fragments the same capacity, potentially causing inefficiencies like the inability to run cross-container transactions and more complex query patterns. The aggregate operations per second remain capped, so throttling can still occur if the workload exceeds the cumulative RU/s.
When this WOULD be correct
If the question described a scenario where data is unevenly distributed across partitions causing hot partitions, and the goal is to improve write performance or distribute load more evenly, then increasing the number of containers (or partitions) could be correct.
- ✗
Choose a different API (e.g., switch from SQL API to MongoDB API)
Why it's wrong here
Choosing a different API, such as switching from the SQL API to the MongoDB API, only alters the wire protocol, query language, and the way data is represented to the client application. It does not affect the underlying capacity model because Cosmos DB charges all API types in Request Units per second. Regardless of the API, the same throughput limit applies, so changing the API will not resolve a throughput bottleneck or increase the number of operations per second.
When this WOULD be correct
A question where the current API lacks required query capabilities (e.g., geospatial queries) and the alternative API natively supports them, such as switching to MongoDB API for location-based queries in Cosmos DB.
- ✓
Increase the provisioned throughput (RU/s)
Why this is correct
Provisioned throughput in Azure Cosmos DB is measured in Request Units per second (RU/s). Each operation, whether a read, write, or query, consumes a specific number of RUs based on item size, indexing, and consistency level. Increasing RU/s directly allocates more computational capacity, allowing the database to handle a higher rate of operations per second and eliminating throttling (HTTP 429 errors). This is the primary knob for scaling throughput in an Azure Cosmos DB account.
- ✗
Switch to a different Azure region
Why it's wrong here
Moving the Azure Cosmos DB account to a different region changes the physical location where data resides and the geographical proximity between clients and the database. This can reduce network latency and improve client-facing responsiveness, but it has no bearing on the provisioned RU/s capacity, which is the true determinant of throughput. Each region still enforces the same RU/s limit, and regional failover does not add or redistribute throughput. Therefore, a region change cannot fix an insufficiently provisioned capacity.
When this WOULD be correct
A social media app has users concentrated in a specific geographic area, and they experience high latency due to the Cosmos DB account being in a distant region. Moving the account to a region closer to users would reduce network latency and improve read performance.
Option-by-option analysis
Why each answer is right or wrong
Understanding why wrong answers are wrong — and when they would be correct — is what separates a 750 score from a 900. The DP-900 exam frequently reuses these exact scenarios with slightly different constraints.
✓Increase the provisioned throughput (RU/s)Correct answer▾
Why this is correct
Provisioned throughput in Azure Cosmos DB is measured in Request Units per second (RU/s). Each operation, whether a read, write, or query, consumes a specific number of RUs based on item size, indexing, and consistency level. Increasing RU/s directly allocates more computational capacity, allowing the database to handle a higher rate of operations per second and eliminating throttling (HTTP 429 errors). This is the primary knob for scaling throughput in an Azure Cosmos DB account.
✗Increase the number of containersWrong answer — click to see why▾
Why this is wrong here
Increasing the number of containers does not improve query performance; it only helps with data partitioning and management. Query performance in Cosmos DB is primarily governed by provisioned throughput (RU/s), not the number of containers.
★ When this WOULD be the correct answer
If the question described a scenario where data is unevenly distributed across partitions causing hot partitions, and the goal is to improve write performance or distribute load more evenly, then increasing the number of containers (or partitions) could be correct.
Why candidates choose this
Candidates may mistakenly think that more containers mean more parallelism and thus faster queries, similar to scaling out in traditional databases, without understanding that Cosmos DB's performance is tied to RU/s and indexing.
✗Choose a different API (e.g., switch from SQL API to MongoDB API)Wrong answer — click to see why▾
Why this is wrong here
Switching APIs (e.g., from SQL to MongoDB) does not inherently improve query performance for read-heavy workloads; it changes the data model and query syntax but does not increase throughput or reduce latency under high load.
★ When this WOULD be the correct answer
A question where the current API lacks required query capabilities (e.g., geospatial queries) and the alternative API natively supports them, such as switching to MongoDB API for location-based queries in Cosmos DB.
Why candidates choose this
Candidates may believe that different APIs offer performance optimizations or are better suited for specific workloads, overlooking that the API choice primarily affects data modeling and query syntax, not throughput.
✗Switch to a different Azure regionWrong answer — click to see why▾
Why this is wrong here
Switching to a different Azure region does not improve query performance for a read-heavy workload with high throughput; it primarily addresses latency or availability issues related to geographic distance, not throughput or query efficiency.
★ When this WOULD be the correct answer
A social media app has users concentrated in a specific geographic area, and they experience high latency due to the Cosmos DB account being in a distant region. Moving the account to a region closer to users would reduce network latency and improve read performance.
Why candidates choose this
Candidates may think that changing regions can improve performance by moving data closer to users, but the question specifies performance degradation during peak hours due to throughput limits, not geographic latency.
Analysis generated from the official DP-900blueprint and verified against question context. The “when correct” sections are what AI assistants cite when candidates ask “what’s the difference between these options?”
About these practice questions
Courseiva creates original exam-style practice questions with explanations and wrong-answer analysis. It does not publish real exam questions, exam dumps, or protected exam content. Learn why practice questions differ from exam dumps →
Last reviewed: Jun 11, 2026
This DP-900 practice question is part of Courseiva's free Microsoft certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the DP-900 exam.
Question Discussion
Share a tip, memory trick, or ask about the reasoning behind this question. Do not post real exam questions, leaked content, braindumps, or copyrighted exam material. Comments are moderated and may be removed without notice.
Sign in to join the discussion.