DP-420 cheat sheet

A one-page reference for the Microsoft Azure Cosmos DB Developer Specialty (DP-420) exam: the format, how the domains are weighted, and the glossary terms for this exam.

Exam at a glance

Vendor
Microsoft
Level
Intermediate
Questions
60
Time
120 min
Mock pass mark
70%
Domains
5
Practice Qs
104
Code
DP-420

Domain weightings

How much of the exam each domain covers. Spend your study time in proportion — the heavier the domain, the more questions you'll see.

Key terms

Azure Cosmos DB
Azure Cosmos DB is Microsoft's globally distributed, multi-model NoSQL database with single-digit-millisecond latency and elastic scale. DP-420 focuses on building cloud-native applications on its API for NoSQL.
Partition Key
A Partition Key is the document property Azure Cosmos DB uses to distribute data across logical partitions for scale. DP-420 makes choosing a high-cardinality, even-distribution partition key the central data-modeling decision.
Logical Partition
A Logical Partition is the set of all items in a container that share the same partition key value, and it has a 20 GB storage limit. DP-420 uses it as the unit of transactional scope and scaling.
Request Unit (RU)
A Request Unit (RU) is the normalized currency of throughput in Azure Cosmos DB, abstracting CPU, memory, and IOPS into one cost per operation. DP-420 expects you to estimate and optimize RU consumption for queries and writes.
Provisioned Throughput
Provisioned Throughput is the model where you reserve a fixed number of Request Units per second, as standard or autoscale, for a database or container. DP-420 contrasts it with serverless for spiky or low-volume workloads.
Autoscale
Autoscale is a provisioned-throughput mode that automatically scales Request Units per second between 10% and 100% of a set maximum based on usage. DP-420 uses it to handle variable traffic without manual intervention.
Consistency Level
A Consistency Level defines the trade-off between data freshness, availability, latency, and RU cost across replicas. DP-420 covers the five levels — strong, bounded staleness, session, consistent prefix, and eventual — and when to use each.
Multi-region Writes
Multi-region Writes let an Azure Cosmos DB account accept writes in more than one region, improving write availability and latency. DP-420 pairs it with a conflict-resolution policy to handle concurrent updates.
Conflict Resolution
Conflict Resolution is the policy that decides which write wins when multi-region writes produce competing updates, using last-writer-wins or a custom procedure. DP-420 requires configuring it when enabling multi-region writes.
Change Feed
The Change Feed is a persistent, ordered record of changes to items in a container that applications can process incrementally. DP-420 uses it with Azure Functions for denormalization, aggregation, referential integrity, and archiving.
Analytical Store
The Analytical Store is a column-oriented copy of container data, automatically synced from the transactional store, for analytics without impacting operational workloads. DP-420 covers reaching it through Microsoft Fabric mirroring, the current recommended path for new analytics projects.
Time to Live (TTL)
Time to Live (TTL) is a setting that automatically deletes items after a specified period, configurable per container and per item. DP-420 uses it to expire transient data without manual cleanup.
ETag
An ETag is a version identifier on each Azure Cosmos DB item used to implement optimistic concurrency control. DP-420 uses it to prevent lost updates by rejecting a write if the item changed since it was read.
Hierarchical Partition Key
A Hierarchical Partition Key (HPK) uses multiple levels of properties as the partition key so that a shared prefix can hold more than the 20 GB per single key value, while each full key path stays within the logical-partition limits. DP-420 uses it for workloads such as multi-tenant designs that would otherwise create hot or oversized partitions.
Indexing Policy
An Indexing Policy defines which item paths Azure Cosmos DB indexes and how, controlling query performance and RU cost. DP-420 covers tuning it — including composite indexes and read-heavy vs write-heavy strategies.