I85PREP
⌘ K
NCM-DB study materials
NCM-DB study resources

NCM-DB practice questions

One question at a time. Build your NCM-DB understanding.

First 50 of 200 questions. Each one lists the choices, the correct answer, and an explanation. The checker at the top of this page saves progress in this browser when it is available.

Continue in the practice tool above to load more questions, 50 at a time.

Question 1Database Architecture, Concepts, and Design

In relational database theory, what does the term cardinality describe when applied to a single relation (table)?

  1. The number of columns (attributes) defined in the relation's schema
  2. The number of rows (tuples) currently stored in the relation Correct answer.
  3. The number of indexes built on the relation's columns
  4. The number of foreign keys that reference the relation

Explanation

Cardinality of a relation refers to the count of tuples, that is, rows in the table. Degree, by contrast, is the number of attributes or columns. These two dimensions are independent: a table can have high cardinality but low degree, or the reverse.

Question 2Database Architecture, Concepts, and Design

A designer places a customer's full address (street, city, postal code) into a single attribute of the Customer relation. Which normalization principle does this most directly violate?

  1. First normal form, because the attribute holds a non-atomic composite value Correct answer.
  2. Second normal form, because a non-key attribute depends on part of a composite key
  3. Third normal form, because a non-key attribute determines another non-key attribute
  4. Boyce-Codd normal form, because every determinant must be a candidate key

Explanation

First normal form requires that every attribute value be atomic, meaning indivisible within the relational model. Storing a composite address in one column creates a repeating-group-like structure that must be decomposed into separate attributes or a related table. The other normal forms address dependency issues, not atomicity.

Question 3Database Architecture, Concepts, and Design

Which key is chosen from among the candidate keys to uniquely identify each row and is declared as the table's primary identifier?

  1. The foreign key
  2. The primary key Correct answer.
  3. The alternate key
  4. The superkey

Explanation

A primary key is the candidate key selected by the designer to serve as the table's unique row identifier. Alternate keys are the candidate keys not chosen, and superkeys are any sets that uniquely identify rows. A foreign key references another table rather than identifying rows in its own table.

Question 4Database Architecture, Concepts, and Design

A database with a single server handles all reads and writes; if that server fails, the application is unavailable. Which architectural property does this design most clearly lack?

  1. Data independence
  2. High availability Correct answer.
  3. Normalization
  4. Referential integrity

Explanation

High availability requires that the system continue serving requests despite component failure, typically through redundancy such as replicas or failover. A single unmirrored server is a single point of failure. The other options describe schema and integrity properties, not fault tolerance.

Question 5Database Architecture, Concepts, and Design

In an entity-relationship model, what distinguishes a weak entity from a strong entity?

  1. A weak entity may only participate in one-to-one relationships
  2. A weak entity's existence depends on an identifying owner entity and it lacks a key of its own Correct answer.
  3. A weak entity cannot have any attributes of its own
  4. A weak entity must be implemented as a view rather than a table

Explanation

A weak entity has no key attribute sufficient to identify its instances independently, so it relies on a partial key combined with the key of an identifying owner entity. Its existence is therefore existence-dependent on that owner. Weak entities can have attributes and participate in various relationship cardinalities.

Question 6Database Architecture, Concepts, and Design

What is the primary purpose of a database index?

  1. To reduce the storage footprint of the table
  2. To guarantee that transactions are serializable
  3. To enforce domain constraints on column values
  4. To provide an alternate access path that speeds up row retrieval Correct answer.

Explanation

An index is a physical data structure, commonly a B-tree, that provides a fast lookup path to rows matching a search condition. It trades additional storage and write overhead for faster reads. Constraints and isolation are enforced by other mechanisms.

Question 7Database Architecture, Concepts, and Design

Which property of the ACID set guarantees that a committed transaction's changes survive a subsequent system crash?

  1. Consistency
  2. Durability Correct answer.
  3. Isolation
  4. Atomicity

Explanation

Durability guarantees that once a transaction commits, its effects are persisted, typically through write-ahead logging and forced log flushes. Atomicity concerns all-or-nothing execution, consistency concerns valid state transitions, and isolation concerns concurrent visibility.

Question 8Database Architecture, Concepts, and Design

A designer creates a separate table for each month of sales data, such as Sales_Jan, Sales_Feb, and Sales_Mar, all with identical columns. Which schema design problem does this introduce?

  1. The design violates third normal form within each individual table
  2. The design creates a partial dependency on a composite key
  3. The design introduces a transitive dependency on the primary key
  4. The design fragments a single logical entity across multiple physical tables Correct answer.

Explanation

Splitting one logical entity into many identically structured tables by period is a form of uncontrolled denormalization that complicates queries, constraints, and reporting. A single table with a period column, optionally partitioned physically, preserves the logical entity. The per-table schemas themselves may still be normalized.

Question 9Database Architecture, Concepts, and Design

Consider a relation Enrollment(StudentID, CourseID, InstructorName) where each course is taught by exactly one instructor. Which normalization step is required?

  1. Remove StudentID to a separate table, because it appears in the composite key
  2. Add a surrogate key without changing any attributes
  3. Remove InstructorName to a separate table keyed by CourseID, because CourseID determines InstructorName Correct answer.
  4. Split the relation by semester into multiple tables

Explanation

The composite key (StudentID, CourseID) determines InstructorName only through CourseID, so InstructorName depends transitively on the key via a non-key attribute. Third normal form requires removing this dependency by placing InstructorName in a table keyed by CourseID. Adding a surrogate key alone does not resolve the dependency.

Question 10Database Architecture, Concepts, and Design

An application must store documents whose fields vary widely from one record to another and are read as complete units. Which data model is most appropriate?

  1. A star schema with a central fact table and fixed dimensions
  2. A strictly normalized relational schema with one column per possible field
  3. A third normal form schema with an entity-attribute-value table for every field
  4. A document-oriented store that keeps each record as a self-describing document Correct answer.

Explanation

Document stores accommodate heterogeneous, self-describing records and are optimized for retrieving a whole document by key. Relational schemas with one column per field or EAV designs impose rigidity and query complexity that do not fit this access pattern. Star schemas are analytical and assume a stable dimensional structure.

Question 11Database Architecture, Concepts, and Design

A fact table records individual sales transactions, and a dimension table stores product details. Which key in the fact table references the dimension?

  1. A surrogate key generated for each fact row
  2. A foreign key pointing to the dimension's primary key Correct answer.
  3. A natural key copied from the source system's invoice number
  4. A composite key formed from all dimension attributes

Explanation

In a star schema, the fact table holds foreign keys that reference the primary keys of surrounding dimension tables. These keys join facts to their descriptive context. Surrogate keys on facts identify fact rows, and natural invoice numbers are business identifiers, not dimension references.

Question 12Database Architecture, Concepts, and Design

A table has a composite primary key of (OrderID, LineNumber). A designer adds a column ProductName that depends only on OrderID. Which normal form does this violate?

  1. Fourth normal form
  2. Second normal form Correct answer.
  3. Third normal form
  4. First normal form

Explanation

Second normal form prohibits a non-key attribute from depending on only part of a composite candidate key. ProductName depends on OrderID alone, a proper subset of the key, which is a partial dependency. First normal form concerns atomicity, and third and fourth normal forms address transitive and multivalued dependencies respectively.

Question 13Database Architecture, Concepts, and Design

A column stores values that are almost always distinct, such as a globally unique identifier. How does this affect the usefulness of a B-tree index on that column for equality lookups?

  1. The index becomes a full table scan because every value is unique
  2. The index only helps range queries and never equality queries
  3. The index is useless because high cardinality prevents efficient tree traversal
  4. The index is highly selective, so equality lookups typically touch very few rows Correct answer.

Explanation

High cardinality means each indexed value matches few rows, so an equality predicate is highly selective and the B-tree can locate the matching entry efficiently. Low-cardinality columns, by contrast, may match many rows and make the index less attractive. Uniqueness does not disable index traversal.

Question 14Database Architecture, Concepts, and Design

Which statement best describes physical data independence?

  1. The database can run on multiple operating systems without recompilation
  2. Applications are unaffected by changes to storage structures such as adding an index or reorganizing files Correct answer.
  3. The physical schema must always match the logical schema exactly
  4. Applications are unaffected by changes to the logical schema such as adding a column

Explanation

Physical data independence means the internal or physical schema, covering storage structures and access paths, can change without requiring changes to the conceptual schema or application queries. Logical data independence concerns changes to the conceptual schema. The two are distinct levels of the three-schema architecture.

Question 15Database Architecture, Concepts, and Design

A many-to-many relationship between Student and Course must be implemented in a relational schema. What is the standard approach?

  1. Add a foreign key to Student referencing Course
  2. Merge both entities into a single table with duplicated columns
  3. Create an associative table with foreign keys to both Student and Course Correct answer.
  4. Add a foreign key to Course referencing Student

Explanation

A many-to-many relationship cannot be represented by a foreign key on either side alone because each row must reference multiple rows on the other side. An associative, or junction, table holds one foreign key per entity and typically forms a composite key. Merging entities would duplicate data and lose independence.

Question 16Database Architecture, Concepts, and Design

In a schema, a foreign key column permits NULL values. What does this allow with respect to referential integrity?

  1. The child row must reference the parent with the smallest key value
  2. The foreign key constraint is silently dropped for the whole table
  3. Rows may exist with no referenced parent, which is permitted unless the column is declared NOT NULL Correct answer.
  4. The parent table can be deleted without any checks

Explanation

Referential integrity constrains non-null foreign key values to match a parent key, but a NULL means no reference exists and is generally allowed when the column is nullable. Declaring the column NOT NULL forces every child row to reference a parent. The constraint itself remains active for non-null values.

Question 17Database Architecture, Concepts, and Design

Which characteristic distinguishes a data warehouse from an operational transaction processing system?

  1. A warehouse is optimized for high-volume short transactions with frequent updates
  2. A warehouse is always implemented on a single server without redundancy
  3. A warehouse enforces stricter row-level locking than an operational system
  4. A warehouse is subject-oriented, integrated, and typically stores historical data for analysis Correct answer.

Explanation

Data warehouses are designed for analytical queries over integrated, subject-oriented historical data, often using denormalized or dimensional schemas. Operational systems favor many small, concurrent update transactions. Locking behavior and server topology are implementation choices, not defining warehouse traits.

Question 18Database Architecture, Concepts, and Design

A designer chooses a surrogate integer key instead of a natural key such as a national identification number for the Customer table. What is the main advantage of this choice?

  1. Surrogate keys eliminate the need for any unique constraint
  2. Surrogate keys reduce the number of columns needed for the primary key to zero
  3. Surrogate keys are stable and independent of business rules that may change Correct answer.
  4. Surrogate keys guarantee that duplicate customers can never be inserted

Explanation

A surrogate key has no business meaning, so changes to business identifiers do not ripple through foreign keys. Natural keys can change format or be reassigned, which complicates references. Surrogate keys do not by themselves prevent duplicate business entities and still require a unique constraint on the natural key.

Question 19Database Architecture, Concepts, and Design

What does the CAP theorem state about a distributed data system?

  1. Partition tolerance is optional and can always be sacrificed
  2. A system can simultaneously guarantee consistency, availability, and partition tolerance
  3. Consistency and availability are the same property under replication
  4. A system can guarantee at most two of consistency, availability, and partition tolerance when a partition occurs Correct answer.

Explanation

The CAP theorem holds that during a network partition a distributed system must choose between consistency and availability; it cannot guarantee all three. Partition tolerance is generally unavoidable in distributed deployments, so designs lean toward CP or AP behavior. Consistency and availability are distinct properties.

Question 20Database Architecture, Concepts, and Design

A schema stores the same customer address in the Orders table and the Shipments table, and updates are applied to only one of them. Which design consequence is most likely?

  1. An improvement in referential integrity because both tables are self-contained
  2. A guaranteed increase in query performance for all workloads
  3. A loss of data integrity due to update anomalies from redundant storage Correct answer.
  4. A violation of first normal form in both tables

Explanation

Storing the same fact in multiple places creates redundancy that leads to update anomalies when only one copy is changed. Normalization removes such redundancy by keeping the address in a single referenced table. The tables may still satisfy first normal form because each attribute value remains atomic.

Question 21Database Architecture, Concepts, and Design

An ER diagram shows a ternary relationship among Supplier, Part, and Project. What does this relationship represent?

  1. A weak entity that depends on all three participants
  2. A hierarchy in which Project is a subtype of Part
  3. A single association whose instances involve one occurrence of each of the three entities Correct answer.
  4. Three separate binary relationships that must be modeled independently

Explanation

A ternary relationship associates instances of three entity types simultaneously, such as recording that a specific supplier supplies a specific part to a specific project. It is not equivalent to three independent binary relationships because the combination matters. It does not imply subtyping or weak-entity status.

Question 22Database Architecture, Concepts, and Design

In a relational schema, a table has no declared primary key and contains duplicate identical rows. Which property is violated?

  1. The requirement that indexes exist on every column
  2. The requirement that all attributes be of the same data type
  3. The requirement that every table have at least one foreign key
  4. The relational model's requirement that tuples be uniquely identifiable within a relation Correct answer.

Explanation

In the formal relational model, a relation is a set of tuples, so duplicate tuples cannot exist and each tuple must be uniquely identifiable. A primary key enforces this uniqueness in practice. There is no requirement that attributes share a type, that foreign keys exist, or that every column be indexed.

Question 23Database Architecture, Concepts, and Design

A relation R(A, B, C) has the functional dependencies A -> B and B -> C. Which statement is correct?

  1. R is in Boyce-Codd normal form because every attribute is functionally determined
  2. R is in third normal form because A is a candidate key
  3. R violates third normal form because C is transitively dependent on the candidate key A Correct answer.
  4. R violates second normal form because B is not a candidate key

Explanation

With A -> B and B -> C, C depends on A only through the non-key attribute B, which is a transitive dependency. Third normal form forbids a non-key attribute determining another non-key attribute, so R must be decomposed. Second normal form is not the issue because the key A is not composite.

Question 24Database Architecture, Concepts, and Design

A designer decomposes a relation to achieve lossless-join dependency preservation. Which statement accurately describes the lossless-join property?

  1. Natural joining the decomposed relations reproduces exactly the original relation without spurious tuples Correct answer.
  2. The decomposed relations must always be joined with an outer join to recover all tuples
  3. The decomposition must reduce the total number of attributes across all relations
  4. The decomposition must produce relations with no overlapping attributes

Explanation

A decomposition is lossless when the natural join of the resulting relations yields precisely the original relation, with no extra tuples or lost tuples. This requires that the shared attributes of the decomposed relations include a key of at least one of them. Overlapping attributes are in fact necessary for a natural join.

Question 25Database Architecture, Concepts, and Design

A query filters on a low-cardinality column, such as a status flag with only three distinct values, and returns roughly one third of all rows. How should a B-tree index on that column be evaluated for this query?

  1. It is guaranteed to be chosen by the optimizer regardless of statistics
  2. It may be less efficient than a sequential scan because low selectivity yields many row lookups Correct answer.
  3. It cannot be used for equality predicates on low-cardinality columns
  4. It will always outperform a sequential scan because it avoids reading the table

Explanation

When a predicate matches a large fraction of rows, following the index to fetch many rows can cost more than a sequential scan. The optimizer uses statistics on value distribution to decide. Indexes remain usable for low-cardinality equality predicates, but they are not automatically beneficial.

Question 26Database Architecture, Concepts, and Design

In a relational schema, a candidate key is defined as a minimal set of attributes whose values uniquely identify each tuple. What does "minimal" mean in this definition?

  1. The key must be declared before any other constraint in the CREATE TABLE statement
  2. No proper subset of the attribute set still uniquely identifies every tuple Correct answer.
  3. The attribute set contains the fewest possible columns of any type in the table
  4. The key values must be stored in sorted order to support efficient lookups

Explanation

Minimality is about irreducibility, not column count or storage order: a candidate key loses its uniqueness guarantee if any attribute is removed. A superset that still uniquely identifies tuples is a superkey but not a candidate key. Minimality is therefore a property of the attribute set relative to the given relation and its functional dependencies.

Question 27Database Architecture, Concepts, and Design

An entity-relationship model contains a weak entity type whose existence depends on an owner entity type. How is the weak entity typically represented in the resulting relational schema?

  1. A view over the owner table that filters rows belonging to the dependent entity
  2. A single column added to the owner table holding a comma-separated list of dependent values
  3. A table whose primary key is the combination of the owner's primary key and the weak entity's partial key Correct answer.
  4. A table with an auto-generated surrogate key and no reference to the owner table

Explanation

A weak entity has no key of its own, so its identifying relationship is mapped by borrowing the owner's primary key. The resulting composite primary key combines the owner key with the weak entity's partial key, which guarantees uniqueness only within one owner. This also enforces the existence dependency through the foreign key to the owner.

Question 28Database Architecture, Concepts, and Design

A relation is in third normal form (3NF) but not in Boyce-Codd normal form (BCNF). Which situation must hold?

  1. The relation contains a composite candidate key and no non-prime attributes at all
  2. Some non-trivial functional dependency has a determinant that is not a superkey, while every determinant is a superkey or a prime attribute Correct answer.
  3. The relation has a partial dependency of a non-prime attribute on part of a candidate key
  4. Every functional dependency has a determinant that is a superkey, but a multivalued dependency is present

Explanation

3NF permits a functional dependency X to Y when X is a superkey or when every attribute of Y is prime. BCNF tightens this by requiring every non-trivial determinant to be a superkey. A relation that is 3NF but not BCNF therefore has a determinant that is not a superkey while its dependent attributes are all prime.

Question 29Database Architecture, Concepts, and Design

A database architect is choosing between a normalized schema and a denormalized schema for a read-heavy reporting workload. Which trade-off does denormalization primarily introduce?

  1. Reduced storage footprint because duplicated columns are compressed automatically
  2. Stronger enforcement of referential integrity through additional foreign keys
  3. Faster reads from fewer joins, at the cost of redundant storage and update anomalies Correct answer.
  4. Guaranteed elimination of phantom reads during concurrent reporting queries

Explanation

Denormalization precomputes or duplicates data so queries need fewer joins, which typically speeds reads. The price is redundant storage and the risk that copies diverge unless update logic maintains them consistently. It does not strengthen integrity constraints or affect isolation-level phenomena such as phantom reads.

Question 30Database Architecture, Concepts, and Design

A university records that each student may enroll in many courses and each course may enroll many students, and that each enrollment carries a grade. In the relational schema, where should the grade attribute be placed?

  1. In the associative table that represents the enrollment relationship, keyed by student and course Correct answer.
  2. In a separate table with no foreign keys, linked to students by name
  3. In the Student table, as a repeating group of grade values
  4. In the Course table, because grades belong to the course offering

Explanation

Attributes of a many-to-many relationship belong to the table that implements that relationship. The enrollment table's composite key of student and course identifies each pairing, so the grade is functionally dependent on the whole key. Placing it in either entity table would create redundancy or violate normalization.

Question 31Database Architecture, Concepts, and Design

A relation R(A, B, C, D) has functional dependencies A -> B and B -> C, with A the only candidate key. Which normal form does R violate?

  1. Boyce-Codd normal form only, while still satisfying third normal form
  2. Second normal form, because C depends on part of the candidate key A
  3. Third normal form, because B determines C while B is not a superkey Correct answer.
  4. First normal form, because attribute C is not atomic

Explanation

The dependency B -> C has a determinant B that is neither a superkey nor a prime attribute, since only A is prime. That is exactly the transitive dependency pattern prohibited by third normal form. Because the violating determinant is not a superkey, the relation also fails BCNF, but the 3NF violation is the more informative answer here.

Question 32Database Architecture, Concepts, and Design

Which statement correctly distinguishes a conceptual data model from a physical data model?

  1. The conceptual model specifies tablespace allocation, while the physical model specifies business rules
  2. The conceptual model is produced only after the physical model has been deployed and tuned
  3. The conceptual model applies only to NoSQL systems, while the physical model applies only to relational systems
  4. The conceptual model describes entities and relationships independent of any DBMS, while the physical model specifies storage structures, indexes, and data types Correct answer.

Explanation

Conceptual modeling captures the business view of entities, attributes, and relationships without committing to a particular product. Physical modeling translates that design into concrete DBMS artifacts such as tables, columns with data types, indexes, partitions, and storage parameters. The two levels serve different audiences and stages of the design lifecycle.

Question 33Database Architecture, Concepts, and Design

An application must support queries that filter on any combination of five optional search attributes over a very large table. Which physical design choice is most appropriate?

  1. Avoid indexes entirely and rely on full table scans for predictable performance
  2. Create a single composite index containing all five attributes in alphabetical order
  3. Create a small set of composite indexes whose leading columns match the most common filter combinations Correct answer.
  4. Create one separate single-column index on each of the five attributes and rely on the optimizer to combine them

Explanation

A composite index can only be used efficiently when the query predicates match its leading columns, so indexes should be designed around the most frequent predicate patterns. Single-column indexes can be combined by the optimizer, but that is often less efficient than a well-chosen composite index for common combinations. Ordering columns alphabetically ignores query patterns and yields a poor leading column.

Question 34Database Architecture, Concepts, and Design

In a relational database, what is the role of a foreign key constraint?

  1. It automatically creates a clustered index on the referencing column
  2. It guarantees that every value in the referencing column is unique across the table
  3. It prevents the referenced table from being updated by any transaction
  4. It ensures that values in the referencing column match a candidate key value in the referenced table or are null where permitted Correct answer.

Explanation

A foreign key enforces referential integrity: each non-null referencing value must correspond to an existing key value in the parent table. The constraint may allow nulls, depending on its definition, and it constrains deletes and updates of the parent through the chosen referential action. It does not by itself impose uniqueness on the child column.

Question 35Database Architecture, Concepts, and Design

A designer must model a relationship in which each employee has exactly one manager and each manager supervises many employees, all within the same Employee entity type. How is this recursive relationship best represented relationally?

  1. Add a ManagerID column to the Employee table that references the Employee table's primary key Correct answer.
  2. Store the manager's name in a text column of the Employee table
  3. Create a junction table with two columns, both referencing the Employee table
  4. Create a separate Manager table containing one row per manager

Explanation

A one-to-many recursive relationship is implemented by a self-referencing foreign key in the same table. The ManagerID column points back to the primary key of Employee, capturing the hierarchy without duplicating entity data. A junction table would be appropriate for a many-to-many recursive relationship instead.

Question 36Database Architecture, Concepts, and Design

A relation is decomposed into two relations to remove redundancy. Which property guarantees that no information is lost when the original relation is reconstructed?

  1. The decomposition places every attribute in exactly one of the resulting relations
  2. The decomposition preserves all functional dependencies in the original relation
  3. The decomposition results in relations that are each in second normal form
  4. The decomposition is a lossless-join decomposition, so the natural join of the parts reproduces exactly the original relation Correct answer.

Explanation

Lossless join is the property that the natural join of the decomposed relations yields precisely the original relation, with no spurious tuples. It holds when the common attributes of the two parts form a superkey of at least one of them. Dependency preservation is a distinct and desirable property but does not by itself prevent tuple loss.

Question 37Database Architecture, Concepts, and Design

Which of the following best describes the purpose of an entity-relationship diagram during database design?

  1. To specify the transaction isolation level used by each application program
  2. To communicate entities, attributes, and relationships at a conceptual level before tables are defined Correct answer.
  3. To record the execution plans generated by the query optimizer
  4. To document the physical file layout and index structures of the deployed database

Explanation

An ER diagram is a conceptual design artifact that captures the things of interest, their properties, and how they relate, using notation such as cardinality and participation constraints. It is intended for communication with stakeholders and as a basis for later logical and physical design. It does not address storage structures, isolation settings, or optimizer plans.

Question 38Database Architecture, Concepts, and Design

A designer is deciding between storing a derived attribute such as OrderTotal, computed from line items, versus computing it at query time. Which consideration most directly favors storing the derived value?

  1. Storage requirements decrease because the total column replaces the line item rows
  2. Read performance improves, but every change to line items must update the stored total to keep it consistent Correct answer.
  3. Referential integrity is enforced automatically on the derived column
  4. The schema moves into a higher normal form by including the derived value

Explanation

Persisting a derived attribute trades consistency risk for read speed: the value must be recomputed or maintained whenever its source data changes. Storing it does not reduce storage, since the line items still exist. It also introduces redundancy, which generally works against normalization rather than toward it.

Question 39Database Architecture, Concepts, and Design

A relation has attributes A, B, C, D with functional dependencies AB -> C and C -> D, and AB is the only candidate key. Which statement about this relation is correct?

  1. It is in second normal form but not third normal form, because C determines D while C is not a superkey Correct answer.
  2. It is in third normal form because AB is a candidate key and no partial dependencies exist
  3. It is in Boyce-Codd normal form because every attribute appears in some dependency
  4. It is in first normal form only because D is not atomic

Explanation

There is no partial dependency, since C and D depend on the whole key AB, so the relation satisfies second normal form. However, the transitive dependency C -> D has a determinant that is neither a superkey nor a prime attribute, which violates third normal form. The presence of a non-superkey determinant also means BCNF is not satisfied.

Question 40Database Architecture, Concepts, and Design

Which statement accurately describes the difference between a superkey and a candidate key in a relation?

  1. Every candidate key is a superkey, but a superkey may contain redundant attributes that are not needed for uniqueness Correct answer.
  2. A superkey applies only to foreign keys, while a candidate key applies only to primary keys
  3. A superkey must be declared in the schema, while a candidate key exists only conceptually
  4. Every superkey is a candidate key, but a candidate key may contain redundant attributes

Explanation

A superkey is any attribute set that uniquely identifies tuples, so it may include extra attributes. A candidate key is a minimal superkey, meaning removal of any attribute destroys uniqueness. The relationship is therefore one of containment: candidate keys form a subset of superkeys.

Question 41Installation, Configuration, and Instance Management

Which Nutanix component runs on each AHV host and provides the local storage, network, and virtualization services that form the distributed platform?

  1. Nutanix Guest Agent
  2. Prism Central
  3. Foundation VM
  4. Controller VM (CVM) Correct answer.

Explanation

Each AHV or ESXi host runs a Controller VM that owns the physical disks and serves storage and other services to the cluster. Prism Central is a management plane, not a per-host data path component.

Question 42Installation, Configuration, and Instance Management

During initial cluster creation with Foundation, what is the recommended network state for the nodes before imaging begins?

  1. Nodes configured with DHCP addresses on a management subnet
  2. All nodes on the same flat Layer 2 network reachable by the Foundation VM Correct answer.
  3. Each node on a different VLAN with routed access
  4. Nodes isolated on separate physical switches with no interconnect

Explanation

Foundation discovers and images nodes over a flat Layer 2 segment so it can broadcast-discover them and push images. Routing or VLAN separation between the Foundation host and the nodes prevents discovery.

Question 43Installation, Configuration, and Instance Management

Which Prism Element feature allows you to apply a consistent set of configuration settings, such as NTP servers and DNS, across multiple clusters from Prism Central?

  1. Categories
  2. Cluster management policies Correct answer.
  3. Projects
  4. Security Policies

Explanation

Prism Central cluster management policies let administrators enforce settings like NTP, DNS, and password policies across registered clusters. Categories and Projects are for VM organization and governance, not host-level configuration enforcement.

Question 44Installation, Configuration, and Instance Management

An administrator needs to add a new node to an existing AHV cluster. Which prerequisite must be satisfied first?

  1. The node must be imaged with the same AOS version and hypervisor as the cluster Correct answer.
  2. The node must be added to Prism Central before cluster expansion
  3. The node must be a different hardware model than existing nodes
  4. The node must run a newer AOS version than the cluster

Explanation

Expansion requires the new node to match the cluster's AOS and hypervisor version so the CVM can join cleanly. Version mismatches are resolved by imaging the node to the correct version before expansion.

Question 45Installation, Configuration, and Instance Management

Which service on the CVM is responsible for handling storage I/O requests from the hypervisor and other CVMs?

  1. Stargate Correct answer.
  2. Zookeeper
  3. Prism
  4. Cassandra

Explanation

Stargate is the data I/O service on the CVM that manages the storage tier and serves read and write requests. Cassandra stores cluster metadata, and Zookeeper handles cluster coordination.

Question 46Installation, Configuration, and Instance Management

Which tool should be used to generate a configuration file that Foundation can consume for unattended node imaging?

  1. Foundation Applet or Foundation Central Correct answer.
  2. acli
  3. Prism Element
  4. ncli

Explanation

Foundation provides an applet (or Foundation Central) to build the imaging configuration, including IP ranges and cluster settings, for unattended deployment. ncli and acli are command-line interfaces for managing an already-deployed cluster.

Question 47Installation, Configuration, and Instance Management

What is the purpose of the Prism Central instance sizing exercise during deployment?

  1. To choose the hypervisor version for the clusters
  2. To determine the number of CVMs required per cluster
  3. To calculate the network bandwidth needed for replication
  4. To select the appropriate vCPU and memory footprint based on the number of managed VMs and clusters Correct answer.

Explanation

Prism Central sizing depends on the scale of managed clusters and VMs, which determines the deployed vCPU and memory profile. It does not affect CVM count or hypervisor version selection.

Question 48Installation, Configuration, and Instance Management

Which CLI command displays the health and status of services running on a CVM?

  1. allssh genesis status Correct answer.
  2. ncli cluster info
  3. cluster status
  4. acli vm.list

Explanation

The genesis service orchestrates CVM services, and running 'allssh genesis status' across CVMs shows the state of each service. ncli cluster info reports cluster-level information, not per-service health.

Question 49Installation, Configuration, and Instance Management

An administrator wants to change the cluster's DNS servers after deployment. Where is this configured in Prism Element?

  1. Settings > Alerts
  2. VM > Categories
  3. Settings > Cluster Details > Name Servers Correct answer.
  4. Hardware > Disks

Explanation

DNS and NTP servers are configured under the cluster details settings in Prism Element. Alert settings and hardware views do not control name resolution.

Question 50Installation, Configuration, and Instance Management

Which statement best describes the role of Prism Central in a multi-cluster environment?

  1. It is required for a single cluster to function
  2. It replaces Prism Element for all local cluster management tasks
  3. It provides centralized management, reporting, and policy enforcement across registered clusters Correct answer.
  4. It stores all VM data for the registered clusters

Explanation

Prism Central is a management plane that aggregates multiple clusters for centralized operations and policies, while Prism Element remains the local cluster manager. It does not store VM data and is optional for a single cluster.