Match each Azure data consistency model to its description.
Drag a concept onto its matching description — or click a concept then click the description.
Reads always see the latest write
Reads may lag behind writes by up to K versions or T time
Consistent reads within a client session
Reads never see out-of-order writes
No ordering guarantee, eventually consistent
Answer choices
Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.
Correct answer & explanation
Strong: Reads always return the most recent write.
Azure Cosmos DB offers five consistency levels: strong, bounded staleness, session, consistent prefix, and eventual. Strong returns the most recent write; bounded staleness allows a bounded lag; session ensures per-session guarantees.
Answer analysis
For each option: why learners choose it and why it is or isn't the right answer here.
Strong: Reads always return the most recent write.
Why this is correct
Strong consistency is the strictest model: once a write is acknowledged, every subsequent read, regardless of replica, must return that value. It provides linearizability because operations appear to execute atomically in a real-time order. This eliminates any visible lag, but in distributed systems it can increase latency because writes must be confirmed by a quorum of replicas before acknowledgement.
Bounded staleness: Reads may lag behind writes by at most a configured time interval or number of versions.
Why this is correct
Bounded staleness deliberately allows reads to return data that is out of date, but only within a configurable maximum — a time interval (for example, 5 seconds) or a maximum number of versions lag. This is a middle-ground consistency guarantee: it gives a predictable upper bound on how stale data can be while permitting better availability and lower latency than strong consistency. Under this model, once the bound is reached, reads cannot continue to return older data; the replica must catch up.
Session: Guarantees monotonic reads, monotonic writes, read your writes, and write follows reads within a session.
Why this is correct
Session consistency scopes guarantees to an individual client session, using a session token to coordinate reads and writes on that session. It ensures monotonic reads (you won't see older data after seeing newer data), monotonic writes, read-your-writes, and write-follows-reads — all within the same session. Outside that session, other clients may see different, eventually consistent data, which makes it a pragmatic default for many applications.
Strong: Reads may lag behind writes by up to 5 seconds.
Why it's wrong here
This statement mistakenly describes bounded staleness rather than strong consistency. Strong consistency has no lag: a read must always reflect the latest acknowledged write on any replica. A delay of up to 5 seconds is an explicit staleness bound, which is exactly what the bounded staleness model permits. Therefore, calling this 'strong' mischaracterizes the zero-lag linearizable guarantee that the strong model provides.
Session: Guarantees that all replicas have the same data at all times.
Why it's wrong here
This statement incorrectly attributes global replica synchronization to session consistency. Strong consistency is the model that guarantees all replicas appear to have the same data at all times, because reads return the latest write. Session consistency only guarantees a consistent view for the duration of a single client session, not across all replicas for all clients. Other clients outside the session may observe older or divergent versions, so the 'all replicas same at all times' property does not hold for session consistency.
Go deeper
Learn chapter
Data Roles and Core Concepts
Key term
Data
Data is raw, unprocessed information, like numbers, words, or measurements, that can be stored, processed, and analyzed by computers.
Key term
Azure Cosmos DB
Azure Cosmos DB is a fully managed, globally distributed NoSQL database service that offers fast reads and writes anywhere in the world with automatic scaling and multiple consistency models.
About these practice questions
Courseiva writes every DP-900 question from scratch — 820 in total, each with an explanation and a wrong-answer breakdown. None are copied from real exams or dumps. Learn why practice questions differ from exam dumps →
Written by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
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.