AZ-305 Design data storage solutions Practice Question
Exhibit
{
"query": "AzureDiagnostics | where ResourceProvider == 'MICROSOFT.DOCUMENTDB' | where Category == 'DataPlaneRequests' | summarize avg(DurationMs) by bin(TimeGenerated, 1h), OperationName | order by avg_DurationMs desc",
"resultSample": "TimeGenerated, OperationName, avg_DurationMs\n2024-01-01 10:00:00, Query, 150\n2024-01-01 11:00:00, Query, 300\n2024-01-01 12:00:00, Query, 450"
}Refer to the exhibit. You run a KQL query against Azure Cosmos DB diagnostics logs. The query shows increasing latency for Query operations over time. Which is the most likely root cause?
⚠ Common exam trap
A common mix-up: candidates confuse throttling (Option C) with latency degradation, but throttling is an immediate rejection (HTTP 429), not a gradual latency increase; the key clue is the 'increasing latency over time' combined with 'increasing RU consumption,' which points to a hot partition, not a capacity issue.
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
✓
A query is consuming increasing RU over time, possibly due to a hot partition
The query latency increase over time, combined with the fact that the query is consuming more Request Units (RU) per execution, strongly indicates a hot partition. In Azure Cosmos DB, a hot partition occurs when a disproportionate amount of traffic hits a single physical partition, causing that partition's RU budget to be exhausted while others remain underutilized. This leads to increased latency for queries targeting that partition, as the partition's resources become saturated.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✓
A query is consuming increasing RU over time, possibly due to a hot partition
Why this is correct
Increasing RU per query commonly signals a hot partition: a single partition key receives disproportionate request volume, forcing that physical partition to consume more throughput and causing cross-partition query scans. As relevant documents accumulate under one key, the query's logical operations grow, which raises RU charge while response time climbs. Monitor per-partition metrics and Query RU/min to confirm.
- ✗
The Cosmos DB account has reached its storage limit
Why it's wrong here
Cosmos DB storage limit is per logical partition (currently 20 GB) and total account storage governed by replication; hitting it yields HTTP 403 (RequestEntityTooLarge) or write failures, not a progressive rise in RU cost for read queries. Existing queries against already-stored data would continue to bill the same RU unless data distribution changes. Storage exhaustion also doesn't increase latency; it blocks writes.
- ✗
The Cosmos DB account is being throttled due to insufficient RUs
Why it's wrong here
Throttling from insufficient RU returns 429 (TooManyRequests) and relies on retry-after logic, so a latency chart would show spikes and retries, not a monotonically increasing RU-per-query trend. While retries add extra RU for retried operations, the underlying query's RU cost is fixed; the real issue is capacity, not query design. Without 429 events, this diagnosis doesn't fit.
- ✗
There is network latency between the client and the Cosmos DB endpoint
Why it's wrong here
Network latency adds a constant baseline to round-trip time between client and endpoint, independent of per-query RU consumption and query complexity. It would affect all operations identically, not progressively degrade over time as data grows. If RU per query is rising, the slowness is server-side query execution rather than transport-level delay.
Go deeper
Related to this question
About these practice questions
This AZ-305 question is part of Courseiva's 795-question bank — original exam-style content with full explanations and wrong-answer analysis, never real exam questions or exam dumps. Learn why practice questions differ from exam dumps →
JA
Written by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
This AZ-305 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 AZ-305 exam.