Question 1 - checkbox, shuffle, partial credit, easy difficulty

Question category: Module: Module 1: Spark, Hortonworks, HDFS, CAP

Select all components that are part of the HDFS distributed file system.

*A: NameNode

Feedback: Correct! NameNode is a critical component of HDFS.

*B: DataNode

Feedback: Correct! DataNode is a critical component of HDFS.

C: JobTracker

Feedback: Incorrect. JobTracker is related to MapReduce, not HDFS.

D: ResourceManager

Feedback: Incorrect. ResourceManager is related to YARN, not HDFS.

*E: Secondary NameNode

Feedback: Correct! Secondary NameNode is a part of HDFS.

F: TaskTracker

Feedback: Incorrect. TaskTracker is related to MapReduce, not HDFS.

Question 2 - checkbox, shuffle, partial credit, easy difficulty

Question category: Module: Module 1: Spark, Hortonworks, HDFS, CAP

Which of the following are features of the Hadoop Distributed File System (HDFS)?

*A: Fault tolerance

Feedback: Correct! HDFS is designed to be highly fault-tolerant, making it reliable for storing large
datasets.

*B: Efficient replication

Feedback: Well done! HDFS efficiently replicates data to ensure fault tolerance and reliability.
*C: Supports various programming languages

Feedback: Good job! HDFS supports multiple programming languages, making it versatile for
developers.

D: Low latency

Feedback: Incorrect. HDFS is optimized for high throughput rather than low latency.

E: Single point of failure

Feedback: Wrong. HDFS is designed to avoid single points of failure through data replication.

Question 3 - checkbox, shuffle, partial credit, easy difficulty

Question category: Module: Module 1: Spark, Hortonworks, HDFS, CAP

Which of the following are characteristics of cloud infrastructure?

*A: Scalability

Feedback: Correct! Scalability is a key characteristic of cloud infrastructure.

*B: Elasticity

Feedback: Correct! Elasticity allows cloud resources to be scaled up or down based on demand.

C: Single-tenancy

Feedback: Incorrect. Cloud infrastructure typically supports multi-tenancy, not single-tenancy.

*D: On-demand self-service

Feedback: Correct! Cloud infrastructure allows users to provision resources on-demand without human
intervention.

E: Fixed pricing

Feedback: Wrong. Cloud services often use a pay-as-you-go pricing model, rather than fixed pricing.

Question 4 - checkbox, shuffle, partial credit, easy difficulty

Question category: Module: Module 1: Spark, Hortonworks, HDFS, CAP

Which of the following frameworks are built on top of Apache Spark?

*A: GraphX

Feedback: Correct! GraphX is a framework built on top of Spark for graph processing.

*B: Hive on Spark

Feedback: Correct! Hive on Spark allows Hive to run on Spark for better performance.

*C: Mllib

Feedback: Correct! Mllib is Spark's machine learning library.

D: Hadoop

Feedback: Incorrect. Hadoop is not built on top of Spark; it is a separate ecosystem.

E: Flume

Feedback: Incorrect. Flume is a distributed service for collecting and transporting log data, not built on
Spark.

Question 5 - checkbox, shuffle, partial credit, easy difficulty

Question category: Module: Module 1: Spark, Hortonworks, HDFS, CAP

Which of the following are key features of Apache Mesos?

*A: Scalability to thousands of nodes

Feedback: Correct! Apache Mesos is designed to scale to thousands of nodes.

*B: Support for various programming languages

Feedback: Correct! Apache Mesos supports frameworks written in various programming languages.

C: Inefficient resource allocation

Feedback: Incorrect. Apache Mesos is known for its efficient resource allocation.

*D: Centralized resource management

Feedback: Correct! Apache Mesos uses a centralized resource management approach.

E: Lack of fault tolerance

Feedback: Incorrect. Apache Mesos provides fault tolerance.

F: Manual scaling

Feedback: Incorrect. Apache Mesos supports automatic scaling.

Question 6 - checkbox, shuffle, partial credit, easy difficulty

Question category: Module: Module 1: Spark, Hortonworks, HDFS, CAP

Which of the following components are part of Hortonworks? Select all that apply.

*A: Apache Hadoop

Feedback: Correct! Apache Hadoop is a key component of Hortonworks.

*B: Zeppelin

Feedback: Correct! Zeppelin is also included in Hortonworks.

C: Docker

Feedback: Incorrect. Docker is not a component of Hortonworks.

*D: Hive

Feedback: Correct! Hive is another component of Hortonworks.

E: PostgreSQL

Feedback: Incorrect. PostgreSQL is not a component of Hortonworks.

Question 7 - checkbox, shuffle, partial credit, easy difficulty

Question category: Module: Module 1: Spark, Hortonworks, HDFS, CAP

Which of the following frameworks are built on top of Spark?

*A: GraphX

Feedback: Correct! GraphX is a graph processing framework built on top of Spark.

*B: Hive on Spark

Feedback: Correct! Hive on Spark allows Hive to run on the Spark execution engine.

*C: Mllib
Feedback: Correct! Mllib is Spark's scalable machine learning library.

D: TensorFlow

Feedback: Incorrect. TensorFlow is a separate machine learning framework developed by Google.

E: Hadoop

Feedback: Incorrect. Hadoop is a different framework for distributed storage and processing.

Question 8 - checkbox, shuffle, partial credit, medium

Question category: Module: Module 1: Spark, Hortonworks, HDFS, CAP

Select all the characteristics that are essential for a robust cloud infrastructure.

*A: Elasticity

Feedback: Correct! Elasticity allows the system to handle varying loads efficiently.

B: Single point of failure

Feedback: Incorrect. A robust cloud infrastructure should avoid single points of failure.

*C: Resource pooling

Feedback: Correct! Resource pooling is essential for optimizing the use of resources.

*D: Scalability

Feedback: Correct! Scalability is crucial for meeting the demands of a growing user base.

E: Manual updates

Feedback: Incorrect. Automated updates are preferred in a robust cloud infrastructure.

Question 9 - text match, easy difficulty

Question category: Module: Module 1: Spark, Hortonworks, HDFS, CAP

What is the programming model used by Google for processing large data sets with a distributed
algorithm on a cluster? Please answer in all lowercase.

*A: mapreduce

Feedback: Correct! MapReduce is the programming model used by Google.

Default Feedback: Incorrect. The correct answer is a programming model developed by Google.

Question 10 - text match, easy difficulty

Question category: Module: Module 1: Spark, Hortonworks, HDFS, CAP

Name a component of Hortonworks used for distributed storage. Please answer in all lowercase.

*A: hdfs

Feedback: Correct! HDFS is used for distributed storage in Hortonworks.

*B: hadoop

Feedback: Correct! Hadoop is a framework that includes HDFS for distributed storage.

Default Feedback: Incorrect. Review Hortonworks components related to distributed storage.

Question 11 - numeric, easy difficulty

Question category: Module: Module 1: Spark, Hortonworks, HDFS, CAP

How many major distributions of cloud computing applications are discussed in this lesson?

*A: 3.0

Feedback: Correct! The lesson discusses three major distributions: Hortonworks, Cloudera, and MapR.

Default Feedback: Incorrect. Consider revisiting the section on major distributions of cloud computing
applications.

Question 12 - text match, easy difficulty

Question category: Module: Module 1: Spark, Hortonworks, HDFS, CAP

What is the term used for managing and scheduling resources in Apache Mesos? Please answer in all
lowercase.

*A: mesos master

Feedback: Correct! The Mesos Master manages and schedules resources across the cluster.

*B: mesos-master

Feedback: Correct! The Mesos Master manages and schedules resources across the cluster.
Default Feedback: Incorrect. The term refers to the central component in Apache Mesos responsible for
managing and scheduling resources.

Question 13 - text match, easy difficulty

Question category: Module: Module 1: Spark, Hortonworks, HDFS, CAP

What is the term used to describe Spark's method of fault tolerance by maintaining the history of
operations that built an RDD? Please answer in all lowercase.

*A: lineage

Feedback: Correct! Spark uses lineage information to provide fault tolerance.

*B: lineageinfo

Feedback: Correct! Spark uses lineage information to provide fault tolerance.

*C: lineageinformation

Feedback: Correct! Spark uses lineage information to provide fault tolerance.

Default Feedback: Incorrect. Review the material on Spark's fault tolerance mechanisms.

Question 14 - text match, easy difficulty

Question category: Module: Module 1: Spark, Hortonworks, HDFS, CAP

Name the MapR tool used for data streaming. Please answer in all lowercase. Please answer in all
lowercase.

*A: mapr streams

Feedback: Correct! MapR Streams is used for data streaming in MapR.

*B: streams

Feedback: Correct! MapR Streams is used for data streaming in MapR.

*C: maprstreams

Feedback: Correct! MapR Streams is used for data streaming in MapR.

Default Feedback: Incorrect. Please review the MapR tools for data streaming.

Question 15 - text match, easy difficulty

Question category: Module: Module 1: Spark, Hortonworks, HDFS, CAP

What is the primary storage system used by Hadoop for large-scale data processing? Please answer in all
lowercase.

*A: hdfs

Feedback: Correct! HDFS is the primary storage system used by Hadoop for large-scale data processing.

*B: hadoop distributed file system

Feedback: Correct! Hadoop Distributed File System (HDFS) is the primary storage system used by
Hadoop for large-scale data processing.

Default Feedback: Incorrect. Please review the primary storage system used by Hadoop for large-scale
data processing.

Question 16 - numeric, easy difficulty

Question category: Module: Module 1: Spark, Hortonworks, HDFS, CAP

How many distinct stages are there in the MapReduce programming model?

*A: 3.0

Feedback: Correct! The MapReduce programming model consists of three distinct stages: the map stage,
the shuffle stage, and the reduce stage.

Default Feedback: Incorrect. Please review the stages of the MapReduce programming model in the
course materials.

Question 17 - numeric, easy difficulty

Question category: Module: Module 1: Spark, Hortonworks, HDFS, CAP

What is the minimum replication factor in HDFS to ensure fault tolerance?

*A: 3.0

Feedback: Correct! The minimum replication factor in HDFS is three to ensure fault tolerance.

Default Feedback: Incorrect. Review the fault tolerance mechanisms in HDFS to find the correct
replication factor.

Question 18 - text match, easy difficulty

Question category: Module: Module 1: Spark, Hortonworks, HDFS, CAP

What term describes the method Spark uses to provide fault tolerance by maintaining a record of the
transformations applied to the data? Please answer in all lowercase.

*A: lineage

Feedback: Correct! Spark uses lineage information to track the transformations applied to data for fault
tolerance.

Default Feedback: Incorrect. Spark's method of maintaining a record of transformations applied to data
is known as lineage.

Question 19 - multiple choice, shuffle, easy difficulty

Question category: Module: Module 1: Spark, Hortonworks, HDFS, CAP

What is the primary purpose of Resilient Distributed Datasets (RDDs) in Apache Spark?

*A: To provide a fault-tolerant collection of elements that can be operated on in parallel.

Feedback: Correct! RDDs are designed to handle data in parallel while providing fault tolerance.

B: To store data in a centralized database for quick access.

Feedback: Not quite. RDDs are distributed across a cluster rather than centralized.

C: To ensure data is only processed once to improve performance.

Feedback: Incorrect. RDDs are about fault tolerance and parallel processing, not single-process
execution.

D: To replicate data across different data centers to avoid data loss.

Feedback: That's not correct. RDDs focus on fault tolerance within a cluster rather than replication
across data centers.

Question 20 - multiple choice, shuffle, easy difficulty

Question category: Module: Module 1: Spark, Hortonworks, HDFS, CAP

What is the main concept used by Apache Spark to handle distributed data processing efficiently?

*A: Resilient Distributed Datasets

Feedback: Correct! Resilient Distributed Datasets (RDDs) are fundamental to Spark's data processing
capabilities.

B: Random Data Distribution

Feedback: Incorrect. Random Data Distribution is not a concept related to Spark. Review the lesson on
RDDs.

C: Reliable Data Dictionaries

Feedback: Incorrect. Reliable Data Dictionaries do not pertain to Spark's core functionalities. Revisit the
RDD concept.

D: Resource Data Deployment

Feedback: Incorrect. Resource Data Deployment is not related to Spark's RDDs. Check the section on
data management in Spark.

Question 21 - multiple choice, shuffle, easy difficulty

Question category: Module: Module 1: Spark, Hortonworks, HDFS, CAP

What is one primary advantage of using HDFS for distributed storage in large-scale data processing?

*A: Fault tolerance through data replication

Feedback: Correct! HDFS is designed to reliably store large amounts of data by replicating it across
multiple nodes, ensuring fault tolerance.

B: Centralized data management

Feedback: Not quite. HDFS is designed to be distributed, which inherently means data management is
decentralized.

C: Real-time data processing

Feedback: Incorrect. While HDFS is efficient for batch processing, it's not inherently designed for real-
time data processing.

D: Support for proprietary programming languages only

Feedback: This isn't correct. HDFS supports various programming languages, not just proprietary ones.

Question 22 - multiple choice, shuffle, easy difficulty

Question category: Module: Module 1: Spark, Hortonworks, HDFS, CAP

What is the primary advantage of using Infrastructure as Code (IaC) in cloud computing environments?

A: Increased manual configuration of resources

Feedback: This is incorrect. IaC is designed to reduce the need for manual configuration, not increase it.

*B: Faster and more reliable deployment of infrastructure

Feedback: Correct! IaC allows for faster deployment and more reliable infrastructure management.

C: Higher costs due to automation

Feedback: This is incorrect. IaC typically reduces costs by automating processes, not increasing them.

D: Reduced flexibility in resource management

Feedback: This is incorrect. IaC actually increases flexibility by allowing for easy adjustments and
scaling.

Question 23 - multiple choice, shuffle, easy difficulty

Question category: Module: Module 1: Spark, Hortonworks, HDFS, CAP

Which of the following systems is specifically designed for Big Data analytics and supports a flexible
and scalable environment?

A: HDFS

Feedback: HDFS is a distributed file system, not a complete system for Big Data analytics.

*B: Spark

Feedback: Correct! Spark is designed for Big Data analytics and supports scalability.

C: SQL

Feedback: SQL is a language used for managing data, not a full system for Big Data analytics.

D: MySQL

Feedback: MySQL is a relational database management system, not a Big Data analytics system.

Question 24 - multiple choice, shuffle, easy difficulty

Question category: Module: Module 1: Spark, Hortonworks, HDFS, CAP

What is one key benefit of using the Hadoop Distributed File System (HDFS)?

*A: Supports efficient data replication and fault tolerance.

Feedback: Correct! HDFS is designed to provide efficient data replication and fault tolerance, ensuring
data reliability and availability.

B: Enables real-time processing of data streams.

Feedback: Not quite. HDFS is primarily designed for batch processing and storage, rather than real-time
data streams.

C: Provides native support for relational database queries.

Feedback: Incorrect. HDFS does not natively support relational database queries. It's optimized for
large-scale data storage.

D: Reduces the need for distributed storage systems.

Feedback: This is incorrect. HDFS is a type of distributed storage system, and it doesn't aim to reduce
their necessity.

Question 25 - multiple choice, shuffle, easy difficulty

Question category: Module: Module 1: Spark, Hortonworks, HDFS, CAP

Which of the following is a key advantage of using Hortonworks tools in cloud computing
environments?

A: Enhanced data security and privacy

Feedback: This is not the primary advantage of Hortonworks tools. Consider the context of distributed
storage and processing.

*B: Efficient distributed storage and processing

Feedback: Correct! Hortonworks tools excel in managing distributed storage and processing, making
them ideal for cloud environments.

C: Simplified user interface design

Feedback: While user interface design is important, it's not a primary advantage of Hortonworks tools in
cloud computing.

D: Cost-effective licensing options

Feedback: Though cost is a consideration, it is not the main advantage of Hortonworks tools.

Question 26 - multiple choice, shuffle, easy difficulty

Question category: Module: Module 1: Spark, Hortonworks, HDFS, CAP

Which of the following is a key component of Hortonworks for distributed storage and processing?

*A: Apache Hadoop

Feedback: Correct! Apache Hadoop is a fundamental component of Hortonworks, providing distributed

storage and processing capabilities.

B: Microsoft Azure

Feedback: Incorrect. Microsoft Azure is a cloud computing platform, not a component of Hortonworks.

C: Google BigQuery

Feedback: Incorrect. Google BigQuery is a data warehouse solution, not associated with Hortonworks.

D: Amazon S3

Feedback: Incorrect. Amazon S3 is a storage service from AWS, not related to Hortonworks
components.

Question 27 - multiple choice, shuffle, easy difficulty

Question category: Module: Module 1: Spark, Hortonworks, HDFS, CAP

Which of the following is a key benefit of using cloud infrastructure in modern applications?

*A: Scalability to handle varying loads

Feedback: Correct! Cloud infrastructure allows applications to scale resources on demand, which is
crucial for handling varying workloads efficiently.

B: Permanent resource allocation

Feedback: Not quite. Cloud resources are typically flexible and can be adjusted based on needs, rather
than being permanently allocated.

C: Manual configuration of all components

Feedback: Incorrect. One of the benefits of cloud infrastructure is automation, which reduces the need
for manual configuration.
D: Increased hardware costs

Feedback: This is not correct. Cloud infrastructure often reduces hardware costs through shared
resources and on-demand pricing models.

Question 28 - multiple choice, shuffle, easy difficulty

Question category: Module: Module 1: Spark, Hortonworks, HDFS, CAP

Which system is specifically designed for real-time data processing?

*A: Spark Streaming

Feedback: Correct! Spark Streaming is designed for real-time data processing.

B: HDFS

Feedback: HDFS is a distributed file system, not specifically for real-time data processing.

C: MapReduce

Feedback: MapReduce is a programming model for processing large datasets but is not focused on real-
time data.

D: Cloudera Manager

Feedback: Cloudera Manager is a management tool, not a system for real-time data processing.

Question 29 - checkbox, shuffle, partial credit, medium

Question category: Module: Module 1: Spark, Hortonworks, HDFS, CAP

Which of the following are true regarding the similarities and differences between YARN and Mesos?

*A: Both YARN and Mesos can manage cluster resources.

Feedback: Correct! Both frameworks are capable of managing resources across clusters efficiently.

*B: YARN is specifically designed for managing Hadoop clusters, whereas Mesos can manage a wide
range of distributed systems.

Feedback: Correct! While YARN is optimized for Hadoop, Mesos is designed to handle various
distributed systems beyond just Hadoop.

*C: Mesos supports Docker containers, but YARN does not.

Feedback: Correct! Mesos has built-in support for containerization technologies like Docker, whereas
YARN traditionally does not.

D: YARN provides better support for multi-tenancy compared to Mesos.

Feedback: This is incorrect. Mesos is generally recognized for its robust multi-tenancy support.

Question 30 - multiple choice, shuffle, medium

Question category: Module: Module 1: Spark, Hortonworks, HDFS, CAP

In Apache Spark, how is fault tolerance achieved to ensure the system can recover from failures?

*A: Fault tolerance is ensured by maintaining lineage information.

Feedback: Correct! By tracking lineage, Spark can reconstruct lost data.

B: Fault tolerance is achieved by using expensive hardware.

Feedback: Not quite. Spark uses lineage information for fault tolerance, not hardware.

C: Fault tolerance is provided by replicating data across multiple nodes.

Feedback: Incorrect. While data replication can offer fault tolerance, Spark specifically uses lineage
information.

D: Fault tolerance is managed by running redundant computations in parallel.

Feedback: This is not correct. Spark uses lineage information to manage fault tolerance, not redundancy.

Question 31 - checkbox, shuffle, partial credit, easy difficulty

Question category: Module: Module 1: Spark, Hortonworks, HDFS, CAP

Which of the following are frameworks built on top of Apache Spark?

*A: GraphX

Feedback: Correct! GraphX is a graph processing framework built on Spark.

B: Hadoop

Feedback: Incorrect. Hadoop is a separate ecosystem, not built on Spark.

*C: Hive on Spark

Feedback: Correct! Hive on Spark is an execution engine for Hive queries on Spark.

D: Spark SQL

Feedback: Incorrect. Spark SQL is a component of Spark but not a separate framework built on top of it.

*E: Mllib

Feedback: Correct! Mllib is a machine learning library built on Spark.

Question 32 - checkbox, shuffle, partial credit, easy difficulty

Question category: Module: Module 1: Spark, Hortonworks, HDFS, CAP

Select the features that are true for the HDFS distributed file system.

*A: Stores large data sets across multiple machines

Feedback: Correct! HDFS is designed to store large datasets across multiple machines.

B: Requires all data to be in a single file

Feedback: Incorrect. HDFS does not require data to be in a single file; it distributes data across multiple
nodes.

*C: Provides fault tolerance through data replication

Feedback: Correct! HDFS provides fault tolerance by replicating data across multiple nodes.

D: Optimized for random read and write access

Feedback: Incorrect. HDFS is optimized for large sequential read and write access, not random access.

*E: Integrates seamlessly with MapReduce for big data processing

Feedback: Correct! HDFS integrates well with MapReduce for processing large datasets.

Question 33 - checkbox, shuffle, partial credit, easy difficulty

Question category: Module: Module 1: Spark, Hortonworks, HDFS, CAP

Which of the following components are part of Hortonworks' offerings? Select all that apply.

*A: Apache Hadoop

Feedback: Correct! Apache Hadoop is a core component of Hortonworks.

*B: Zeppelin

Feedback: Correct! Zeppelin is included in Hortonworks' offerings.

C: Microsoft Azure

Feedback: Incorrect. Microsoft Azure is a separate cloud service provider, not a component of
Hortonworks.

*D: Pig

Feedback: Correct! Pig is part of Hortonworks' toolset for data processing.

E: Google Cloud Functions

Feedback: Incorrect. Google Cloud Functions is a serverless computing service, not part of
Hortonworks.

*F: YARN

Feedback: Correct! YARN is a resource management platform included in Hortonworks.

Question 34 - checkbox, shuffle, partial credit, medium

Question category: Module: Module 1: Spark, Hortonworks, HDFS, CAP

Which of the following are responsibilities of YARN in a Hadoop ecosystem?

*A: Resource allocation

Feedback: Correct! YARN handles resource allocation among applications running in a cluster.

*B: Application scheduling

Feedback: Correct! YARN acts as a job scheduler for applications, ensuring efficient task execution.

C: Data replication

Feedback: Not quite. Data replication is primarily handled by HDFS, not YARN.

D: Security management

Feedback: Incorrect. While security is important, YARN is not responsible for managing security within
the Hadoop ecosystem.

E: Fault tolerance
Feedback: Incorrect. Fault tolerance is primarily managed by HDFS through data replication, not
YARN.

Question 35 - checkbox, shuffle, partial credit, medium

Question category: Module: Module 1: Spark, Hortonworks, HDFS, CAP

Select all the major systems used for data analysis.

*A: Spark

Feedback: Correct! Spark is a major system used for data analysis.

*B: Hortonworks

Feedback: Correct! Hortonworks is a major system for data analysis.

*C: Cloudera

Feedback: Correct! Cloudera is a major system for data analysis.

*D: MapR

Feedback: Correct! MapR is a major system for data analysis.

E: HDFS

Feedback: HDFS is part of the Hadoop ecosystem but is not a standalone system for data analysis.

F: Oracle

Feedback: Oracle is a database management system, not specifically designed for Big Data analysis.

Question 36 - checkbox, shuffle, partial credit, easy difficulty

Question category: Module: Module 1: Spark, Hortonworks, HDFS, CAP

Which of the following tools and functionalities are part of the Hortonworks distribution?

*A: YARN

Feedback: Correct! YARN is a cluster management component in Hortonworks.

*B: Zeppelin

Feedback: Correct! Zeppelin is an interactive data analytics tool included in Hortonworks.

C: Microsoft Power BI

Feedback: Incorrect. Microsoft Power BI is a business analytics tool, not part of Hortonworks.

*D: Hive

Feedback: Correct! Hive is a data warehouse software that facilitates reading, writing, and managing
large datasets, part of Hortonworks.

E: Apache Beam

Feedback: Incorrect. Apache Beam is a unified model for defining both batch and streaming data-
parallel processing pipelines, not directly part of Hortonworks.

Question 37 - checkbox, shuffle, partial credit, easy difficulty

Question category: Module: Module 1: Spark, Hortonworks, HDFS, CAP

Which of the following frameworks are built on top of Apache Spark?

*A: GraphX provides APIs for graphs and graph-parallel computation

Feedback: Correct! GraphX is designed for graph processing.

*B: Hive on Spark allows querying of large datasets using SQL

Feedback: Correct! Hive on Spark enables SQL-like queries over large datasets.

*C: Mllib is used for machine learning algorithms

Feedback: Correct! Mllib is Spark's library for machine learning.

D: SparkStream handles real-time streaming data

Feedback: Incorrect. While SparkStream is part of Spark, it is not a framework built on top of Spark like
GraphX, Hive on Spark, and Mllib.

E: MapReduce runs iterative algorithms efficiently

Feedback: Incorrect. MapReduce is known for its limitations in handling iterative algorithms, unlike
Spark.

Question 38 - text match, easy difficulty

Question category: Module: Module 1: Spark, Hortonworks, HDFS, CAP

What is the name of the programming model used with HDFS for processing large datasets? Please
answer in all lowercase.

*A: mapreduce

Feedback: Correct! MapReduce is the programming model used with HDFS for processing large
datasets.

Default Feedback: Revisit the lesson on HDFS and its associated programming models for processing
data.

Question 39 - text match, easy difficulty

Question category: Module: Module 1: Spark, Hortonworks, HDFS, CAP

What is the programming model used by Hadoop for processing large data sets? Please answer in all
lowercase.

*A: mapreduce

Feedback: Correct! MapReduce is the programming model used by Hadoop.

Default Feedback: Review the components of Hadoop and the programming model it uses for data
processing.

Question 40 - text match, easy difficulty

Question category: Module: Module 1: Spark, Hortonworks, HDFS, CAP

What is the major distribution of cloud computing applications that is known for its integration with
Apache Hadoop and other big data tools? Please answer in all lowercase.

*A: hortonworks

Feedback: Correct! Hortonworks is well-known for its integration with Apache Hadoop and other big
data tools.

B: cloudera

Feedback: Incorrect. While Cloudera is a major distribution, it is not the one known specifically for
integration with Apache Hadoop and other big data tools.

C: mapr

Feedback: Incorrect. MapR is another distribution but is not primarily known for its integration with
Apache Hadoop.
Default Feedback: Incorrect. Think about the distribution that is specifically recognized for its
integration with Apache Hadoop.

Question 41 - text match, easy difficulty

Question category: Module: Module 1: Spark, Hortonworks, HDFS, CAP

What is the primary purpose of Spark's lineage information? Please answer in all lowercase.

*A: faulttolerance

Feedback: Correct! Lineage information helps in rebuilding lost data, ensuring fault tolerance.

*B: fault-tolerance

Feedback: Correct! Lineage information aids in providing fault tolerance by enabling data recovery.

Default Feedback: Incorrect. Consider how Spark maintains data integrity and supports recovery in
cluster environments.

Question 42 - checkbox, shuffle, partial credit, hard

Question category: Module: Module 3: Streaming Systems

Which of the following statements correctly differentiate between parallel data processing paths and
fault tolerance in Lambda and Kappa Architecture?

*A: Lambda Architecture supports both real-time and batch processing.

Feedback: Correct! Lambda Architecture is designed to handle both real-time and batch processing.

*B: Kappa Architecture simplifies the design by only having a single stream processing path.

Feedback: Correct! Kappa Architecture eliminates the batch layer and uses a single stream processing
path.

C: Lambda Architecture eliminates the need for batch processing by using a single stream processing
path.

Feedback: Incorrect. Lambda Architecture includes both batch and real-time processing.

D: Kappa Architecture relies on batch processing for fault tolerance.

Feedback: Incorrect. Kappa Architecture relies on stream processing for both real-time processing and
fault tolerance.
E: Lambda and Kappa Architecture provide the same approach to fault tolerance.

Feedback: Incorrect. Lambda Architecture uses both batch and real-time processing for fault tolerance,
while Kappa Architecture relies solely on stream processing.

Question 43 - checkbox, shuffle, partial credit, medium

Question category: Module: Module 3: Streaming Systems

Which of the following are key requirements for a stream processing framework?

*A: Low latency processing

Feedback: Correct! Low latency processing is a key requirement for a stream processing framework.

*B: High throughput

Feedback: Correct! High throughput is essential for processing large volumes of data in real-time.

C: Frequent disk writes

Feedback: Incorrect. Frequent disk writes are not typically a key requirement for a stream processing
framework.

*D: Scalability

Feedback: Correct! Scalability is crucial to handle varying data loads efficiently.

E: Complex event query capabilities

Feedback: Incorrect. While useful, complex event query capabilities are not a key requirement for all
stream processing frameworks.

Question 44 - checkbox, shuffle, partial credit, easy difficulty

Question category: Module: Module 3: Streaming Systems

Which of the following are components of a Storm topology?

*A: Spout

Feedback: Correct! Spouts are one of the fundamental components of a Storm topology.

*B: Bolt

Feedback: Correct! Bolts process the tuples in a Storm topology.

C: Stream

Feedback: Incorrect. While streams are sequences of tuples in Storm, they are not considered
components of a topology.

D: Nimbus

Feedback: Incorrect. Nimbus is the master node in a Storm cluster, not a component of a topology.

E: Supervisor

Feedback: Incorrect. Supervisor nodes manage workers in a Storm cluster, but are not components of a
topology.

Question 45 - checkbox, shuffle, partial credit, easy difficulty

Question category: Module: Module 3: Streaming Systems

Which of the following are important motivations for distributed stream processing?

*A: Scalability to handle large volumes of data

Feedback: Correct! Scalability is a key motivation for distributed stream processing as it allows the
system to handle large volumes of data efficiently.

*B: Low latency to process and analyze data in real-time

Feedback: Correct! Low latency is crucial for distributed stream processing to provide real-time insights
and responses.

C: Simplifying data storage and retrieval

Feedback: Incorrect. Simplifying data storage and retrieval is not a primary motivation for distributed
stream processing. Think about the real-time processing aspects.

*D: Fault tolerance to ensure continuous data processing

Feedback: Correct! Fault tolerance is essential for distributed stream processing to maintain continuous
data processing despite failures.

E: Reducing the need for data encryption

Feedback: Incorrect. Reducing the need for data encryption is not a motivation for distributed stream
processing. Focus on the processing capabilities and requirements.

Question 46 - checkbox, shuffle, partial credit, medium

Question category: Module: Module 3: Streaming Systems

Which of the following are challenges in real-time big data stream processing?

*A: High throughput

Feedback: Correct! High throughput is essential for processing large volumes of data in real-time.

*B: Fault tolerance

Feedback: Correct! Fault tolerance is crucial to ensure continuous data processing even in case of
failures.

C: Data consistency

Feedback: Incorrect. While important, it is not typically considered a primary challenge in real-time
processing.

*D: Resource management

Feedback: Correct! Efficiently managing resources is vital for handling real-time data streams.

E: Data warehousing

Feedback: Incorrect. Data warehousing is more related to batch processing than real-time stream
processing.

Question 47 - checkbox, shuffle, partial credit, easy difficulty

Question category: Module: Module 3: Streaming Systems

Select the characteristics that define cloud computing.

*A: On-demand self-service

Feedback: Correct. On-demand self-service is a key characteristic of cloud computing, allowing users to
provision resources as needed automatically.

*B: Broad network access

Feedback: Correct. Broad network access is essential as it ensures that services are available over the
network and accessed through standard mechanisms.

*C: Scalability and elasticity

Feedback: Correct. Scalability and elasticity allow cloud resources to be scaled up or down based on
demand.

D: Fixed resource allocation

Feedback: Incorrect. Fixed resource allocation is not a characteristic of cloud computing; cloud
resources are dynamically allocated as needed.

E: High capital investment

Feedback: Incorrect. High capital investment is not associated with cloud computing; it usually reduces
capital expenditure by using a pay-as-you-go model.

Question 48 - text match, easy difficulty

Question category: Module: Module 3: Streaming Systems

Which programming language is used alongside Java in the implementation of the Storm scheduler?
Please answer in all lowercase.

*A: clojure

Feedback: Correct! Clojure is used alongside Java in the implementation of the Storm scheduler.

Default Feedback: Incorrect. Please review the implementation details of the Storm scheduler and try
again.

Question 49 - numeric, easy difficulty

Question category: Module: Module 3: Streaming Systems

What is the typical latency (in milliseconds) goal for real-time stream processing systems?

*A: 100.0

Feedback: Correct! The typical latency goal for real-time stream processing systems is around 100
milliseconds.

Default Feedback: Incorrect. Ensure you review the latency goals for real-time stream processing
systems.

Question 50 - text match, easy difficulty

Question category: Module: Module 3: Streaming Systems

What term describes processing data as it is generated, rather than in batches? Please answer in all
lowercase.

*A: streaming

Feedback: Correct! Streaming describes processing data as it is generated.

*B: stream

Feedback: Correct! Stream is also an acceptable term for processing data as it is generated.

Default Feedback: Incorrect. The correct term describes processing data as it is generated, rather than in
batches.

Question 51 - text match, easy difficulty

Question category: Module: Module 3: Streaming Systems

Which architecture is designed for processing large-scale data in both real-time and batch modes? Please
answer in all lowercase.

*A: lambda

Feedback: Correct! The Lambda architecture processes data using both real-time and batch processing
methods.

*B: lambda architecture

Feedback: Correct! The Lambda architecture processes data using both real-time and batch processing
methods.

Default Feedback: Incorrect. Revisit the concepts of Lambda and Kappa architectures for processing
large-scale data.

Question 52 - numeric, easy difficulty

Question category: Module: Module 3: Streaming Systems

In what year did Yahoo begin moving to real-time data processing?

*A: 2011.0

Feedback: Correct! Yahoo started its move towards real-time data processing in 2011.

Default Feedback: Incorrect. Review the timeline and reasons behind Yahoo's transition to real-time
data processing.
Question 53 - text match, easy difficulty

Question category: Module: Module 3: Streaming Systems

What is the term for distributing workloads across multiple computing resources to ensure no single
resource is overwhelmed? Please answer in all lowercase.

*A: loadbalancing

Feedback: Correct! Load balancing distributes workloads evenly across multiple resources.

*B: load-balancing

Feedback: Correct! Load balancing distributes workloads evenly across multiple resources.

*C: load_balance

Feedback: Correct! Load balancing distributes workloads evenly across multiple resources.

Default Feedback: Incorrect. The term refers to the process of distributing workloads across multiple
computing resources to prevent any single resource from becoming overwhelmed.

Question 54 - numeric, easy difficulty

Question category: Module: Module 3: Streaming Systems

What is the recommended number of nodes for setting up a small Apache Storm cluster for development
purposes?

*A: 3.0

Feedback: Correct! A small Apache Storm cluster for development typically consists of 3 nodes.

Default Feedback: Incorrect. Consider revisiting the environment setup guidelines for Apache Storm.

Question 55 - numeric, easy difficulty

Question category: Module: Module 3: Streaming Systems

What is the default number of Acker tasks in a Storm topology if not otherwise specified?

*A: 1.0

Feedback: Correct! The default number of Acker tasks in a Storm topology is 1.

Default Feedback: Incorrect. You might want to revisit the section discussing the configuration of Acker
tasks in Storm topologies.

Question 56 - text match, easy difficulty

Question category: Module: Module 3: Streaming Systems

What is the term used to describe the ability of Spark Streaming to maintain state information across
batches? Please answer in all lowercase.

*A: stateful

Feedback: Correct! Spark Streaming can maintain state information across batches through stateful
processing.

*B: statefulprocessing

Feedback: Correct! Spark Streaming can maintain state information across batches through stateful
processing.

Default Feedback: Revisit the concept of stateful stream processing in Spark Streaming.

Question 57 - text match, easy difficulty

Question category: Module: Module 3: Streaming Systems

In the Trident framework, what is the term for ensuring that each tuple is processed exactly once? Please
answer in all lowercase.

*A: exactlyonce

Feedback: Correct! The Trident framework ensures that each tuple is processed exactly once, known as
'exactly once' processing.

*B: exactly-once

Feedback: Correct! The Trident framework ensures that each tuple is processed exactly once, known as
'exactly-once' processing.

*C: exactly_once

Feedback: Correct! The Trident framework ensures that each tuple is processed exactly once, known as
'exactly_once' processing.

Default Feedback: Incorrect. Please review the concept of 'exactly once' processing in the Trident
framework.
Question 58 - numeric, medium

Question category: Module: Module 3: Streaming Systems

Evaluate the performance of a data processing tool that can process a dataset in 12.5 seconds. What is
the speedup factor if another tool can process the same dataset in 7.5 seconds?

*A: 1.67

Feedback: Good job! The speedup factor is calculated by dividing the original time by the new time
(12.5 / 7.5).

Default Feedback: Revisit how speedup factors are calculated by dividing the original execution time by
the new execution time.

Question 59 - multiple choice, shuffle, easy difficulty

Question category: Module: Module 3: Streaming Systems

What is one of the primary goals of Apache Storm as a real-time streaming system?

*A: To process data streams with low latency

Feedback: Correct! Apache Storm is designed to process streams of data with minimal delay, making it
suitable for real-time applications.

B: To store large datasets efficiently

Feedback: Not quite. While data storage is important, Apache Storm's primary focus is on real-time
processing, not storage.

C: To manage distributed databases

Feedback: This option is incorrect. Apache Storm is primarily used for real-time data processing, not for
managing databases.

D: To simplify the creation of machine learning models

Feedback: Incorrect. Apache Storm is not specifically designed for machine learning model
development.

Question 60 - multiple choice, shuffle, easy difficulty

Question category: Module: Module 3: Streaming Systems

What is the role of the IScheduler interface in the Storm scheduling framework?
*A: It determines the allocation of resources among different tasks.

Feedback: Correct! The IScheduler interface plays a crucial role in resource allocation within Storm.

B: It is responsible for compiling Java files from Thrift definitions.

Feedback: Not quite. The IScheduler interface is about scheduling tasks, not compiling files.

C: It manages the interplay between Clojure and Java in Storm.

Feedback: Close, but the IScheduler focuses on scheduling, not managing language interoperability.

D: It provides methods for scaling big data systems.

Feedback: While scaling is important, the IScheduler is primarily about scheduling tasks, not directly
providing scaling methods.

Question 61 - multiple choice, shuffle, easy difficulty

Question category: Module: Module 3: Streaming Systems

What is the main purpose of Acker tasks in Apache Storm's processing framework?

*A: To track the progress of tuples and ensure message processing guarantees.

Feedback: Correct! Acker tasks are responsible for tracking the lineage of tuples and ensuring that they
are processed accurately.

B: To provide a user interface for monitoring topologies.

Feedback: Incorrect. Acker tasks do not provide a user interface.

C: To store the state of a running topology using persistent storage.

Feedback: Incorrect. Persistent storage classes are used for state storage, not Acker tasks.

D: To distribute tuples evenly across different spouts.

Feedback: Incorrect. Acker tasks are not used for tuple distribution across spouts.

Question 62 - multiple choice, shuffle, easy difficulty

Question category: Module: Module 3: Streaming Systems

When deploying cloud applications, which of the following strategies can help in improving the
application's availability?
*A: Implementing automatic scaling mechanisms.

Feedback: Correct! Automatic scaling allows the application to handle varying loads efficiently.

B: Relying solely on a single server architecture.

Feedback: Single server architecture can be a point of failure. Consider revisiting distributed systems
concepts.

C: Ignoring load balancing configurations.

Feedback: Load balancing is crucial for distributing traffic evenly. Avoid this oversight.

D: Focusing exclusively on application aesthetics.

Feedback: While aesthetics are important, they do not directly impact availability. Reflect on the core
functionalities.

Question 63 - multiple choice, shuffle, easy difficulty

Question category: Module: Module 3: Streaming Systems

Which cloud service model focuses primarily on providing virtual machines and other resources as a
service?

*A: Infrastructure as a Service (IaaS)

Feedback: Good understanding of cloud services! IaaS provides virtualized computing resources over
the internet.

B: Platform as a Service (PaaS)

Feedback: PaaS focuses on providing platforms for application development, not directly on virtual
machines.

C: Software as a Service (SaaS)

Feedback: SaaS delivers software over the internet without focusing on virtual machines.

D: Function as a Service (FaaS)

Feedback: FaaS involves executing code in response to events, not managing virtual machines.

Question 64 - multiple choice, shuffle, easy difficulty

Question category: Module: Module 3: Streaming Systems

What is a primary challenge when building stream processing systems?

*A: Handling data in real-time without delays

Feedback: Well done! Real-time processing is critical in stream processing systems to ensure timely
insights.

B: Storing large volumes of static data

Feedback: Stream processing focuses on real-time analysis rather than static data storage. Consider
revisiting the key challenges of stream processing.

C: Converting batch data to stream data

Feedback: While converting data formats can be a task, it's not a primary challenge in stream processing
systems. Think about the core requirements for handling data streams.

D: Ensuring data is serialized properly for storage

Feedback: Serialization is more relevant to data storage and less about real-time stream challenges.
Consider the main focus of stream processing systems.

Question 65 - multiple choice, shuffle, easy difficulty

Question category: Module: Module 3: Streaming Systems

What is the primary function of a Bolt in an Apache Storm topology?

A: To emit streams of data into the topology.

Feedback: This describes the function of a Spout, not a Bolt. Revisit the components of a Storm
topology.

*B: To process incoming streams and pass them to the next component.

Feedback: Correct! Bolts are responsible for processing streams in a topology.

C: To manage the cluster resources required for the topology.

Feedback: Cluster resource management is not the primary function of a Bolt. Look into Nimbus and
Supervisors for this role.

D: To define the data schemas used within the topology.

Feedback: Data schema definition is not the primary role of a Bolt. Consider revisiting the task
breakdown of Storm components.
Question 66 - multiple choice, shuffle, easy difficulty

Question category: Module: Module 3: Streaming Systems

What is the primary challenge of building stream processing systems?

*A: Handling data in motion efficiently

Feedback: Correct! Efficiently handling data in motion is the main challenge in stream processing
systems.

B: Storing large volumes of data permanently

Feedback: Storing data permanently is a challenge for batch processing, not stream processing.

C: Ensuring data privacy and security

Feedback: While important, data privacy and security are not the primary challenges specific to stream
processing.

D: Designing user-friendly interfaces

Feedback: Designing interfaces is not a primary challenge in stream processing systems.

Question 67 - multiple choice, shuffle, medium

Question category: Module: Module 3: Streaming Systems

How does Trident ensure exactly once processing in the context of Apache Storm?

*A: Trident ensures exactly once processing through checkpointing and transactional spouts.

Feedback: Correct! Trident uses checkpointing and transactional spouts to ensure exactly once
processing.

B: Trident ensures exactly once processing by using tuple timeouts and retries.

Feedback: Incorrect. Tuple timeouts and retries are not mechanisms used by Trident for exactly once
processing.

C: Trident ensures exactly once processing by leveraging parallel processing and batch updates.

Feedback: Incorrect. Parallel processing and batch updates are not directly related to exactly once
processing in Trident.

D: Trident ensures exactly once processing by using stateless spouts and non-transactional states.
Feedback: Incorrect. Stateless spouts and non-transactional states do not provide exactly once
processing guarantees.

Question 68 - multiple choice, shuffle, medium

Question category: Module: Module 3: Streaming Systems

What is the primary advantage of using stateful stream processing in Spark Streaming compared to
traditional systems like Storm?

*A: Allows storing and updating state information across batches.

Feedback: Correct! Stateful stream processing in Spark Streaming enables storing and updating state
information, which is crucial for certain applications.

B: Reduces the complexity of the code significantly.

Feedback: Not quite. While stateful processing might simplify some tasks, the primary advantage is the
ability to store and update state information.

C: Increases the processing speed for all types of data.

Feedback: Incorrect. The primary advantage is not necessarily about speed, but about managing state
information effectively.

D: Ensures no data loss in event of failures.

Feedback: This is partially true, but the main advantage of stateful processing is related to state
management rather than fault tolerance.

Question 69 - multiple choice, shuffle, medium

Question category: Module: Module 3: Streaming Systems

What impact does batch size have on system performance and latency in Spark Streaming?

*A: Batch size affects both system performance and latency in Spark Streaming.

Feedback: Correct! Batch size impacts how data is processed and can influence both performance and
latency.

B: Batch size only affects the memory usage of Spark Streaming.

Feedback: Not quite. While batch size does affect memory usage, it also impacts performance and
latency.
C: Batch size has no impact on latency in Spark Streaming.

Feedback: Incorrect. Batch size does have an impact on latency, as well as on performance.

D: Batch size only influences the data ingestion rate in Spark Streaming.

Feedback: That's not correct. Batch size impacts more than just the data ingestion rate.

Question 70 - multiple choice, shuffle, medium

Question category: Module: Module 3: Streaming Systems

What role does the IScheduler interface play in the Apache Storm scheduling framework?

A: It determines the order of task execution across the cluster.

Feedback: The order of task execution is crucial, but the IScheduler focuses more on task placement
rather than order.

*B: It manages resource allocation and task placement on the cluster nodes.

Feedback: Correct! The IScheduler interface is responsible for resource allocation and task placement in
Apache Storm.

C: It handles error detection and recovery mechanisms in the cluster.

Feedback: While error detection and recovery are important, they are not the primary responsibility of
the IScheduler interface.

D: It optimizes data serialization and deserialization processes.

Feedback: Data serialization and deserialization are handled elsewhere, not by the IScheduler interface.

Question 71 - checkbox, shuffle, partial credit, medium

Question category: Module: Module 3: Streaming Systems

Select all practices that are commonly used to improve the performance and reliability of cloud
infrastructure.

*A: Load balancing

Feedback: Load balancing is a key practice in cloud infrastructure for distributing workloads evenly.

*B: Vertical scaling

Feedback: Vertical scaling involves adding more power (CPU, RAM) to an existing machine and is
relevant to cloud infrastructure.

C: Function chaining

Feedback: Function chaining is more relevant to serverless architectures rather than general cloud
infrastructure.

*D: Data sharding

Feedback: Data sharding is commonly used in database management to improve performance, especially
in distributed systems.

E: Manual server configuration

Feedback: Manual server configuration is less common in cloud environments which typically leverage
automation.

Question 72 - checkbox, shuffle, partial credit, medium

Question category: Module: Module 3: Streaming Systems

Which of the following are essential components of a 'Stream Processing' framework?

*A: Real-time data ingestion

Feedback: Correct! Real-time data ingestion is crucial for stream processing.

B: Batch processing capabilities

Feedback: Batch processing is not typically part of a stream processing framework.

*C: Low-latency processing

Feedback: Correct! Low-latency processing is essential for stream processing.

*D: Scalable architecture

Feedback: Correct! A scalable architecture is important for handling variable data loads in stream
processing.

E: Complex user interfaces

Feedback: Complex user interfaces are not essential for stream processing frameworks.

Question 73 - checkbox, shuffle, partial credit, hard

Question category: Module: Module 3: Streaming Systems

Differentiate between the parallel data processing paths and fault tolerance in Lambda and Kappa
Architecture. Select all statements that apply.

*A: Kappa Architecture provides real-time processing with a unified stream processing path.

Feedback: Correct! Kappa Architecture is known for its unified approach to real-time stream processing.

*B: Lambda Architecture requires separate batch and speed layers for fault tolerance.

Feedback: Correct! Lambda Architecture uses separate layers to handle both batch and real-time data
processing.

C: In Lambda Architecture, both batch and real-time data are processed together in a single path.

Feedback: Not quite. Lambda Architecture separates batch and real-time processing to optimize
performance.

D: Kappa Architecture cannot handle real-time data processing.

Feedback: That's incorrect. Kappa Architecture is specifically designed for real-time processing.

E: Lambda Architecture is more suited for systems that need immediate processing of data.

Feedback: While Lambda Architecture can process data in real-time, it is not solely designed for
immediate processing.

Question 74 - checkbox, shuffle, partial credit, medium

Question category: Module: Module 3: Streaming Systems

Select the correct statements regarding the role of Ackers in Storm and the performance impact of
enabling spout support for replay.

*A: Ackers help in identifying failed tuples by tracking tuple lineage.

Feedback: Correct! Ackers track the lineage of tuples to determine their successful or failed processing.

*B: Enabling spout support for replay can decrease performance due to additional overhead.

Feedback: Correct! Spout replay adds overhead, which can impact performance.

C: Ackers eliminate the need for spout replay in Storm.

Feedback: Incorrect. Ackers do not eliminate the need for spout replay; they are complementary
components.

D: Ackers improve performance by reducing the number of emitted tuples.

Feedback: Incorrect. Ackers track tuples but do not directly reduce the number of emitted tuples.

E: Ackers and spout replay are unrelated components in Storm.

Feedback: Incorrect. Ackers and spout replay work together to ensure reliable message processing.

Question 75 - checkbox, shuffle, partial credit, medium

Question category: Module: Module 3: Streaming Systems

Which of the following are true regarding Storm's message processing guarantees using Tuple Trees,
Anchoring, and Spout Replay?

*A: Tuple Trees track the lineage of tuples throughout the topology.

Feedback: Correct! Tuple Trees are used to manage and track the flow of tuples.

*B: Anchoring ensures that tuples can be replayed if they fail to process.

Feedback: Correct! Anchoring is crucial for message replay in case of failures.

C: Spout Replay allows for at-most-once processing guarantees.

Feedback: Incorrect. Spout Replay is used for at-least-once or exactly-once processing guarantees, not
at-most-once.

D: Anchoring eliminates the need for Acker tasks.

Feedback: Incorrect. Anchoring works in conjunction with Acker tasks, not as a replacement.

E: Tuple Trees automatically provide exactly-once processing without additional configuration.

Feedback: Incorrect. Exactly-once processing requires additional configuration and mechanisms beyond
Tuple Trees.

Question 76 - checkbox, shuffle, partial credit, easy difficulty

Question category: Module: Module 3: Streaming Systems

Which of the following are key components of a Storm topology?

*A: Spouts

Feedback: Correct! Spouts are responsible for emitting streams into the system.

*B: Bolts

Feedback: Correct! Bolts process and transform the data within a topology.

C: Nodes

Feedback: Incorrect. In Storm, the term 'nodes' is not used as a specific component of a topology.

*D: Streams

Feedback: Correct! Streams are the core data unit in Storm, representing the flow of data between spouts
and bolts.

E: Clusters

Feedback: Incorrect. While Storm runs on clusters, 'clusters' are not a specific component of a topology.

Question 77 - checkbox, shuffle, partial credit, hard

Question category: Module: Module 3: Streaming Systems

Which of the following tools are suitable for different stages of the data processing pipeline based on
their features and performance benchmarks?

*A: Apache Storm for real-time processing.

Feedback: Correct! Apache Storm is suitable for real-time data processing due to its low latency.

*B: Kafka for data storage and message brokering.

Feedback: Correct! Kafka is excellent for durable data storage and message brokering.

C: Spark Streaming for batch processing only.

Feedback: Incorrect. Spark Streaming is primarily used for stream processing, not just batch processing.

*D: NiFi for visual data flow management.

Feedback: Correct! NiFi provides a robust solution for visual data flow management and control.

E: Lambda Architecture for reducing data redundancy.

Feedback: Incorrect. Lambda Architecture is more about balancing speed and accuracy in data
processing rather than data redundancy reduction.

Question 78 - checkbox, shuffle, partial credit, easy difficulty

Question category: Module: Module 3: Streaming Systems

Which of the following are essential components for cloud infrastructure security? Select all that apply.

*A: Data encryption at rest and in transit.

Feedback: Correct! Data encryption is vital to protect data against unauthorized access.

*B: Regular security audits and assessments.

Feedback: Correct! Regular audits help identify vulnerabilities and improve security posture.

C: Using hard-coded credentials in applications.

Feedback: Hard-coded credentials pose a security risk. Revisit secure coding practices.

*D: Implementing multi-factor authentication.

Feedback: Correct! Multi-factor authentication adds an extra layer of security.

E: Disabling all network firewalls for easier access.

Feedback: Disabling firewalls can expose systems to attacks. Consider network security best practices.

Question 79 - checkbox, shuffle, partial credit, medium

Question category: Module: Module 3: Streaming Systems

Which of the following are important when organizing a real-time Big Data problem in a 'Stream
Processing' framework?

*A: Low-latency processing

Feedback: Correct! Low-latency is crucial for timely insights in stream processing.

B: Batch processing compatibility

Feedback: Batch processing is not typically associated with real-time stream processing frameworks.
Consider the nature of real-time data.

*C: Scalability
Feedback: Correct! Scalability ensures that the system can handle varying data loads efficiently.

D: Fixed data schemas

Feedback: Stream processing often requires flexible schemas due to dynamic data. Think about the
adaptability needed in stream environments.

*E: Fault tolerance

Feedback: Correct! Fault tolerance is vital to ensure system reliability in case of failures.

Question 80 - text match, easy difficulty

Question category: Module: Module 3: Streaming Systems

In cloud computing, what is the term for processing data as it arrives rather than in batches? Please
answer in all lowercase.

*A: streaming

Feedback: Great job! Streaming is key to handling continuous data flow efficiently.

Default Feedback: Consider reviewing the differences between processing data in real-time and in
batches.

Question 81 - text match, easy difficulty

Question category: Module: Module 3: Streaming Systems

What is the term for processing streams of data in real-time within data centers? Please answer in all
lowercase.

*A: streaming

Feedback: Correct! Streaming is the term for processing data in real-time.

*B: streamprocessing

Feedback: Correct! Streamprocessing is an accepted term for this process.

Default Feedback: Consider the continuous flow of data needing immediate processing and revisit the
concept.

Question 82 - checkbox, shuffle, partial credit, easy difficulty

Question category: Module: Module 2: Large Scale Data Storage

Select all key features of Kafka.

*A: Scalability

Feedback: Correct! Kafka is highly scalable, making it suitable for large-scale data applications.

*B: Handling large amounts of data

Feedback: Correct! Kafka is designed to efficiently handle large volumes of data.

C: Centralized architecture

Feedback: Incorrect. Kafka uses a distributed architecture, not a centralized one.

*D: Real-time data processing

Feedback: Correct! Kafka is widely used for real-time data processing.

E: Manual data partitioning

Feedback: Incorrect. Kafka automatically handles data partitioning.

Question 83 - checkbox, shuffle, partial credit, easy difficulty

Question category: Module: Module 2: Large Scale Data Storage

Which of the following are advanced data structures supported by Redis?

*A: Sets

Feedback: Correct! Sets are indeed supported by Redis.

*B: Sorted sets

Feedback: Correct! Sorted sets are indeed supported by Redis.

*C: Hash sets

Feedback: Correct! Hash sets are indeed supported by Redis.

D: Queues

Feedback: Incorrect. Queues are not advanced data structures supported by Redis.

E: Trees
Feedback: Incorrect. Trees are not advanced data structures supported by Redis.

Question 84 - checkbox, shuffle, partial credit, easy difficulty

Question category: Module: Module 2: Large Scale Data Storage

Which of the following are characteristics of the BASE model in distributed systems?

*A: Basic Availability

Feedback: Correct! Basic Availability is one of the characteristics of the BASE model.

B: Strong Consistency

Feedback: Incorrect. Strong Consistency is a characteristic of the ACID model, not BASE.

*C: Soft state

Feedback: Correct! Soft state is a characteristic of the BASE model.

*D: Eventual consistency

Feedback: Correct! Eventual consistency is a key characteristic of the BASE model.

E: Immediate Consistency

Feedback: Incorrect. Immediate Consistency is a characteristic of the ACID model.

Question 85 - checkbox, shuffle, partial credit, medium

Question category: Module: Module 2: Large Scale Data Storage

Which of the following are challenges of storing and managing large amounts of data in distributed
systems?

*A: Latency issues

Feedback: Correct! Latency issues are a significant challenge in distributed systems.

*B: Data corruption

Feedback: Correct! Data corruption can occur and is a challenge in distributed systems.

*C: Single point of failure

Feedback: Correct! Avoiding single points of failure is crucial in distributed systems.

D: High costs of hardware

Feedback: Incorrect. While costs can be a consideration, they are not a primary challenge in managing
data in distributed systems.

E: Insufficient encryption

Feedback: Incorrect. Insufficient encryption relates to security concerns, not the direct challenges of
managing data in distributed systems.

Question 86 - checkbox, shuffle, partial credit, easy difficulty

Question category: Module: Module 2: Large Scale Data Storage

Which of the following are main data types in Spark SQL?

*A: Dataset

Feedback: Correct! Dataset is a main data type in Spark SQL.

*B: DataFrame

Feedback: Correct! DataFrame is a main data type in Spark SQL.

C: RDD

Feedback: Incorrect. RDD is a core data structure in Spark but not a main data type in Spark SQL.

D: Tuple

Feedback: Incorrect. Tuple is not a main data type in Spark SQL.

E: Map

Feedback: Incorrect. Map is not a main data type in Spark SQL.

Question 87 - checkbox, shuffle, partial credit, easy difficulty

Question category: Module: Module 2: Large Scale Data Storage

Which of the following are benefits of using Infrastructure as Code (IaC) in cloud computing?

*A: Consistency in infrastructure setups

Feedback: Correct! IaC ensures that infrastructure setups are consistent across different environments.
*B: Automated infrastructure provisioning

Feedback: Right! IaC allows for automated and repeatable infrastructure provisioning.

C: Increased manual intervention

Feedback: Incorrect. IaC aims to reduce the need for manual intervention, not increase it.

*D: Enhanced scalability

Feedback: Correct! IaC helps in achieving better scalability by allowing infrastructure changes through
code.

E: Higher operational costs

Feedback: No. One of the benefits of IaC is reducing operational costs, not increasing them.

Question 88 - text match, easy difficulty

Question category: Module: Module 2: Large Scale Data Storage

What is the primary use of Kafka in a big data environment? Please answer in all lowercase.

*A: streaming

Feedback: Correct! Kafka is primarily used for real-time data streaming in big data environments.

*B: messaging

Feedback: Correct! Kafka can also be referred to as a messaging system in big data environments.

Default Feedback: Incorrect. Kafka is primarily used for real-time data streaming or messaging in big
data environments.

Question 89 - text match, easy difficulty

Question category: Module: Module 2: Large Scale Data Storage

In the Paxos algorithm, which role is responsible for proposing values? Please answer in all lowercase.

*A: proposer

Feedback: Correct! The proposer is responsible for proposing values in the Paxos algorithm.

*B: proposers
Feedback: Correct! The proposer is responsible for proposing values in the Paxos algorithm.

Default Feedback: Incorrect. Refer to the lesson on the Paxos algorithm to understand the roles of
different agents.

Question 90 - text match, easy difficulty

Question category: Module: Module 2: Large Scale Data Storage

What is the name of the distributed file system developed by Google? Please answer in all lowercase.

*A: gfs

Feedback: Correct! The distributed file system developed by Google is called GFS (Google File
System).

*B: googlefs

Feedback: Correct! The distributed file system developed by Google is also known as GoogleFS.

Default Feedback: Incorrect. Please review the distributed file systems discussed in the lesson.

Question 91 - text match, easy difficulty

Question category: Module: Module 2: Large Scale Data Storage

What is the primary type of data storage used by Redis? Please answer in all lowercase.

*A: in-memory

Feedback: Correct! Redis primarily uses in-memory data storage.

*B: memory

Feedback: Correct! Redis primarily uses in-memory data storage.

Default Feedback: Incorrect. Please review the key features and benefits of Redis.

Question 92 - text match, easy difficulty

Question category: Module: Module 2: Large Scale Data Storage

What is the acronym for the set of properties that guarantee database transactions are processed reliably?
Please answer in all lowercase.

*A: acid
Feedback: Correct! ACID stands for Atomicity, Consistency, Isolation, Durability.

B: acids

Feedback: Incorrect. The correct acronym is ACID without the 's'.

C: acidic

Feedback: Incorrect. The correct acronym is ACID, not acidic.

Default Feedback: Incorrect. Please review the properties that guarantee database transactions are
processed reliably.

Question 93 - text match, easy difficulty

Question category: Module: Module 2: Large Scale Data Storage

What is the term for creating multiple resources that can handle requests and work as a single system in
cloud computing? Please answer in all lowercase.

*A: clustering

Feedback: Correct! Clustering involves creating multiple resources that can work together as a single
system.

*B: cluster

Feedback: Correct! Clustering involves creating multiple resources that can work together as a single
system.

Default Feedback: Incorrect. Review the concept of creating systems that can handle requests
collectively.

Question 94 - numeric, medium

Question category: Module: Module 2: Large Scale Data Storage

If a traditional MySQL database retrieves data in 200 milliseconds, within what range would you expect
Cassandra to retrieve the same data to demonstrate its efficiency?

*A: [50, 150]

Feedback: Good job! This range shows Cassandra's improved efficiency over MySQL.

Default Feedback: Incorrect. Consider Cassandra's efficiency improvements over traditional databases.
Question 95 - multiple choice, shuffle, easy difficulty

Question category: Module: Module 2: Large Scale Data Storage

What is one of the main advantages of Apache Cassandra's ring structure in comparison with traditional
databases like MySQL?

*A: Improved fault tolerance and horizontal scalability

Feedback: Correct! Cassandra's ring structure allows it to scale horizontally and provides improved fault
tolerance.

B: Better integration with cloud services

Feedback: Not quite. While Cassandra can integrate with cloud services, this is not a primary benefit of
its ring structure.

C: Reduced data redundancy

Feedback: Incorrect. The ring structure is not primarily about reducing data redundancy.

D: Simplified data query language

Feedback: That's not correct. The ring structure is unrelated to the complexity of the query language.

Question 96 - multiple choice, shuffle, easy difficulty

Question category: Module: Module 2: Large Scale Data Storage

What is one of the primary uses of Apache Kafka in big data applications?

*A: Kafka is primarily used for real-time data streaming.

Feedback: Correct! Kafka is indeed used for real-time data streaming due to its high throughput and low
latency.

B: Kafka is a type of relational database.

Feedback: This is incorrect. Kafka is not a database; it's a messaging system.

C: Kafka is used mainly for batch processing of data.

Feedback: Incorrect. While Kafka can be used in batch processing contexts, it is designed for real-time
data streaming.

D: Kafka is a programming language.

Feedback: Incorrect. Kafka is not a programming language but a distributed event streaming platform.

Question 97 - checkbox, shuffle, partial credit, easy difficulty

Question category: Module: Module 2: Large Scale Data Storage

Select the characteristics of BASE (Basically Available, Soft state, Eventually consistent) model in
distributed systems.

A: Provides strong consistency guarantees.

Feedback: BASE systems do not provide strong consistency; they offer eventual consistency.

*B: Allows for partial failures without total system shutdown.

Feedback: Correct! BASE systems are designed to handle partial failures gracefully.

C: Offers immediate consistency after every transaction.

Feedback: BASE systems do not offer immediate consistency; they provide eventual consistency over
time.

*D: Sacrifices consistency for availability and partition tolerance.

Feedback: Correct! BASE focuses on availability and partition tolerance, often at the expense of
immediate consistency.

Question 98 - multiple choice, shuffle, easy difficulty

Question category: Module: Module 2: Large Scale Data Storage

Which cloud service model provides virtualized computing resources over the internet?

*A: Infrastructure as a Service (IaaS)

Feedback: Correct! IaaS provides virtualized computing resources over the internet.

B: Platform as a Service (PaaS)

Feedback: Not quite, PaaS provides a platform allowing customers to develop, run, and manage
applications without dealing with infrastructure complexities.

C: Software as a Service (SaaS)

Feedback: Incorrect. SaaS delivers software applications over the internet, on a subscription basis.
D: Network as a Service (NaaS)

Feedback: No, this is a different cloud service model that provides network services over the internet.

Question 99 - multiple choice, shuffle, easy difficulty

Question category: Module: Module 2: Large Scale Data Storage

What is one of the key features of HBase that makes it suitable for managing large-scale data?

*A: Column-oriented storage

Feedback: Correct! HBase's column-oriented storage allows for efficient read and write operations,
especially with sparse data.

B: Row-oriented storage

Feedback: Not quite. While row-oriented storage is common in many databases, HBase uses a column-
oriented approach.

C: Supports SQL natively

Feedback: This is not correct. HBase does not natively support SQL, but it can be integrated with tools
like Phoenix for SQL compatibility.

D: Built-in data compression tools

Feedback: Incorrect. HBase relies on HDFS for storage and does not have built-in compression tools.

Question 100 - multiple choice, shuffle, easy difficulty

Question category: Module: Module 2: Large Scale Data Storage

What is the primary advantage of using a distributed file system in handling large amounts of data?

A: It allows for the centralized storage of data.

Feedback: Distributed file systems are designed to avoid centralized storage to enhance reliability and
accessibility.

*B: It provides a global file namespace for easier data access.

Feedback: Correct! A global file namespace simplifies data access across distributed systems.

C: It ensures data is always stored on a single server for quick retrieval.

Feedback: Single server storage is not a characteristic of distributed file systems.

D: It automatically compresses data to save space.

Feedback: While compression may be used, it is not the primary advantage of distributed file systems.

Question 101 - multiple choice, shuffle, easy difficulty

Question category: Module: Module 2: Large Scale Data Storage

Which cloud service model provides virtualized computing resources over the internet?

*A: Infrastructure as a Service (IaaS)

Feedback: Correct! IaaS provides virtualized computing resources over the internet.

B: Software as a Service (SaaS)

Feedback: Incorrect. SaaS delivers software applications over the internet, not infrastructure.

C: Platform as a Service (PaaS)

Feedback: Incorrect. PaaS provides a platform allowing customers to develop, run, and manage
applications.

D: Backend as a Service (BaaS)

Feedback: Incorrect. BaaS provides web and mobile app developers with a way to connect their
applications to backend cloud storage and APIs exposed by backend applications.

Question 102 - multiple choice, shuffle, easy difficulty

Question category: Module: Module 2: Large Scale Data Storage

What is a primary advantage of using Kafka in big data applications?

A: Kafka ensures data consistency across distributed systems.

Feedback: While Kafka does maintain order within partitions, ensuring data consistency is not its
primary advantage in big data applications. Think about other aspects that make Kafka suitable for
handling large data volumes.

*B: Kafka supports real-time data processing with low latency.

Feedback: Correct! Kafka's ability to process data in real-time with low latency is a significant
advantage in big data applications.
C: Kafka provides an advanced security framework for data protection.

Feedback: Security is important, but Kafka's primary strength in big data scenarios is not its security
features. Consider its data handling capabilities.

D: Kafka allows seamless integration with various cloud service providers.

Feedback: While integration is a feature, Kafka's key advantage in big data applications lies elsewhere.
Reflect on Kafka's core functionality in data processing.

Question 103 - multiple choice, shuffle, easy difficulty

Question category: Module: Module 2: Large Scale Data Storage

What is the main purpose of a distributed file system in handling large amounts of data?

*A: To provide high availability and fault tolerance by replicating data across multiple nodes.

Feedback: That's correct! Distributed file systems are designed to ensure high availability and fault
tolerance by replicating data across multiple nodes, making them ideal for handling large amounts of
data.

B: To enhance data retrieval speed by storing data on a single central server.

Feedback: Not quite. While data retrieval speed can be important, a single central server does not
provide the fault tolerance and scalability offered by distributed systems.

C: To reduce the cost of data storage by minimizing the usage of storage devices.

Feedback: This is incorrect. Distributed file systems focus more on scalability and reliability rather than
reducing storage costs.

D: To allow simultaneous data processing by preventing data redundancy.

Feedback: That's not right. Simultaneous data processing in distributed systems often involves data
replication, which may increase redundancy but enhances reliability and performance.

Question 104 - multiple choice, shuffle, easy difficulty

Question category: Module: Module 2: Large Scale Data Storage

Which of the following best describes the trade-off between consistency and speed in systems using
eventual consistency?

*A: Systems prioritize speed over immediate consistency.

Feedback: Correct! Eventual consistency allows systems to prioritize speed, accepting that consistency
will be achieved eventually.

B: Systems prioritize immediate consistency at the expense of speed.

Feedback: Not quite. Eventual consistency accepts temporary inconsistencies to improve speed.

C: Systems achieve both immediate consistency and high speed simultaneously.

Feedback: This is not correct. Balancing consistency and speed often involves a trade-off.

D: Systems are neither consistent nor fast.

Feedback: Incorrect. The system aims for eventual consistency while maintaining operational speed.

Question 105 - multiple choice, shuffle, medium

Question category: Module: Module 2: Large Scale Data Storage

Which of the following best describes the main challenge of achieving consistency in large-scale
distributed data storage systems?

A: Ensuring immediate data availability across all nodes.

Feedback: Immediate availability often conflicts with consistency guarantees due to network delays and
partitioning issues.

*B: Balancing consistency, availability, and partition tolerance in the presence of network failures.

Feedback: Correct! This refers to the CAP theorem, which highlights the trade-offs between
consistency, availability, and partition tolerance.

C: Increasing storage capacity without affecting performance.

Feedback: While important, this isn't directly related to consistency challenges.

D: Implementing a unified database schema across all nodes.

Feedback: A unified schema can help with data management but doesn't address the core consistency
challenge in distributed systems.

Question 106 - multiple choice, shuffle, medium

Question category: Module: Module 2: Large Scale Data Storage

Which of the following best describes the trade-off involved in achieving consistency in large-scale data
storage systems?

*A: Ensuring consistency often requires compromises on availability and partition tolerance.

Feedback: Correct! This describes the trade-offs in the CAP theorem where consistency, availability,
and partition tolerance cannot all be fully achieved at the same time.

B: Achieving consistency eliminates the need for redundancy.

Feedback: Not quite. Consistency does not eliminate the need for redundancy; it addresses the accuracy
of the data across the system.

C: Consistency ensures data is always available regardless of network failures.

Feedback: Incorrect. Consistency focuses on the correctness of the data, not its availability during
network issues.

D: Consistency is achieved when all users can access data simultaneously.

Feedback: This option mixes up consistency with availability. Consistency ensures correct data, not
simultaneous access for all users.

Question 107 - multiple choice, shuffle, medium

Question category: Module: Module 2: Large Scale Data Storage

Which of the following describes a key benefit of the Paxos algorithm in achieving eventual consistency
in cloud environments?

A: Ensures immediate consistency across all nodes

Feedback: This is incorrect. Paxos aims at eventual consistency, not immediate.

*B: Facilitates consensus among distributed systems

Feedback: Correct! Paxos helps achieve consensus in distributed systems, crucial for eventual
consistency.

C: Reduces latency significantly in data processing

Feedback: While important, reducing latency is not the primary focus of Paxos.

D: Eliminates the need for data replication

Feedback: This is incorrect. Paxos does not eliminate the need for data replication.
Question 108 - multiple choice, shuffle, medium

Question category: Module: Module 2: Large Scale Data Storage

Which of the following describes how HBase manages large-scale data?

A: HBase utilizes a flat file storage system for data management.

Feedback: Not quite. Consider how HBase ensures efficient access and management of large datasets.

*B: HBase employs a three-layer scheme for data location and retrieval.

Feedback: Correct! HBase uses a three-layer scheme which is crucial for efficient data management and
retrieval.

C: HBase relies solely on in-memory storage for its operations.

Feedback: This answer isn't accurate. Think about the role of persistent storage in HBase.

D: HBase uses a centralized database system for data storage.

Feedback: Incorrect. Remember that HBase is designed to be a distributed system.

Question 109 - multiple choice, shuffle, medium

Question category: Module: Module 2: Large Scale Data Storage

What is a key design feature of Apache Cassandra's data storage mechanism?

A: Distributed Hash Table (DHT)

Feedback: The concept of a ring is central to Cassandra's distributed architecture, but it is not called a
Distributed Hash Table.

*B: Ring structure

Feedback: Correct! The ring structure is a fundamental aspect of Apache Cassandra's design, allowing
for efficient data distribution.

C: Master-slave replication model

Feedback: Cassandra uses a peer-to-peer replication model, not master-slave.

D: Hierarchical data storage

Feedback: Cassandra's data storage model is based on a keyspace-column family model, rather than a
hierarchy.

Question 110 - checkbox, shuffle, partial credit, medium

Question category: Module: Module 2: Large Scale Data Storage

Select the characteristics associated with BASE as opposed to ACID in distributed systems.

*A: Eventual consistency

Feedback: Correct! BASE systems aim for eventual consistency rather than immediate consistency.

B: Immediate consistency

Feedback: This is incorrect. Immediate consistency is a characteristic of ACID systems.

*C: High availability

Feedback: Correct! BASE systems often prioritize availability over immediate consistency.

D: Strong data integrity

Feedback: This is incorrect. ACID systems focus on maintaining strong data integrity.

*E: Flexibility

Feedback: Correct! BASE systems offer more flexibility compared to the rigid structure of ACID.

F: Strict transaction completion

Feedback: This is incorrect. Strict transaction completion is a feature of ACID systems.

Question 111 - checkbox, shuffle, partial credit, medium

Question category: Module: Module 2: Large Scale Data Storage

What are the features and benefits of using Spark SQL for structured data processing?

*A: Provides support for both batch and stream processing.

Feedback: Correct! Spark SQL's versatility includes support for both batch and stream processing.

B: Lacks integration capabilities with popular BI tools.

Feedback: Incorrect. Consider how Spark SQL might integrate with data visualization and BI tools.
*C: Offers a unified interface for data processing across different sources.

Feedback: Correct! Spark SQL offers a unified interface, making data processing more seamless across
various sources.

D: Restricts the use of SQL queries for data manipulation.

Feedback: Not quite. Think about how Spark SQL enhances SQL queries in the context of big data.

*E: Optimizes query execution through Catalyst optimizer.

Feedback: Correct! The Catalyst optimizer is one of the key features that enhances query execution in
Spark SQL.

Question 112 - checkbox, shuffle, partial credit, easy difficulty

Question category: Module: Module 2: Large Scale Data Storage

Which of the following are characteristics of the MapReduce programming model?

*A: Automatic parallelization and distribution of tasks.

Feedback: Correct! MapReduce automatically parallelizes and distributes tasks across a cluster, which
simplifies coding and execution.

B: Requires manual intervention for task scheduling.

Feedback: Incorrect. MapReduce abstracts task scheduling, relieving developers from manually
managing it.

C: Supports incremental processing of data streams.

Feedback: Incorrect. While MapReduce can handle large datasets, it is not primarily designed for
incremental processing of data streams.

*D: Handles fault tolerance by re-executing failed tasks.

Feedback: That's right! MapReduce handles fault tolerance by re-executing tasks that fail, ensuring the
robustness of processing large datasets.

E: Optimized for real-time data processing.

Feedback: Incorrect. MapReduce is more suited for batch processing rather than real-time data
processing.

Question 113 - checkbox, shuffle, partial credit, medium

Question category: Module: Module 2: Large Scale Data Storage

Which of the following are key design features and components of Apache Cassandra?

*A: Ring structure

Feedback: Correct! The ring structure is a fundamental part of Cassandra's architecture.

*B: Replication strategies

Feedback: Correct! Replication strategies are crucial for data durability and availability in Cassandra.

C: Foreign key constraints

Feedback: Incorrect. Cassandra does not use foreign key constraints as traditional RDBMS do.

*D: Data storage components

Feedback: Correct! Data storage components are integral to how Cassandra manages and stores data.

E: Triggers for stored procedures

Feedback: Incorrect. Cassandra does not support triggers for stored procedures, unlike some traditional
databases.

Question 114 - checkbox, shuffle, partial credit, easy difficulty

Question category: Module: Module 2: Large Scale Data Storage

Which of the following are benefits of using the MapReduce programming model?

*A: Easy distribution of code across a cluster

Feedback: Correct! MapReduce simplifies the distribution of code across multiple nodes.

B: Immediate real-time data processing

Feedback: MapReduce is not designed for real-time processing; it's more suited for batch processing.

*C: Efficient handling of parallelism in computations

Feedback: Correct! MapReduce is excellent at managing parallel tasks efficiently.

D: Complete elimination of data redundancy

Feedback: Data redundancy is not inherently addressed by MapReduce.

*E: Simplifies large-scale computation

Feedback: Correct! MapReduce effectively simplifies computations on large datasets.

Question 115 - checkbox, shuffle, partial credit, medium

Question category: Module: Module 2: Large Scale Data Storage

Which of the following statements are true about Spark SQL? Select all that apply.

*A: Spark SQL enables querying of structured data using SQL syntax.

Feedback: Correct! Spark SQL supports SQL syntax for querying structured data.

B: Spark SQL only supports DataFrames, not Datasets.

Feedback: Incorrect. Spark SQL supports both DataFrames and Datasets.

C: Spark SQL can be used for real-time data processing.

Feedback: Not quite. While Spark Streaming can handle real-time data, traditional Spark SQL is used
for batch processing.

*D: Spark SQL integrates with Hive for accessing tables.

Feedback: Correct! Spark SQL can integrate with Hive to access tables and use Hive's query language.

E: Spark SQL is limited to working with small data sets.

Feedback: Incorrect. Spark SQL is designed to handle large-scale data processing.

Question 116 - checkbox, shuffle, partial credit, medium

Question category: Module: Module 2: Large Scale Data Storage

Which of the following characteristics are associated with BASE in distributed systems? Select all that
apply.

*A: Basically Available

Feedback: Correct! BASE stands for Basically Available, Soft state, Eventually consistent.

B: Acid Transactions

Feedback: Incorrect. ACID, not BASE, is associated with acid transactions.

*C: Soft state

Feedback: Correct! BASE allows for a soft state, meaning the state may change over time.

*D: Eventual consistency

Feedback: Correct! BASE systems are eventually consistent, which means updates will propagate to all
nodes eventually.

E: Strong Consistency

Feedback: Incorrect. BASE favors eventual consistency, unlike ACID's strong consistency.

Question 117 - checkbox, shuffle, partial credit, medium

Question category: Module: Module 2: Large Scale Data Storage

In the context of distributed systems, which of the following are challenges associated with storing and
managing large amounts of data?

*A: Data consistency across nodes

Feedback: Correct! Ensuring data consistency across distributed nodes is a significant challenge in
distributed systems.

B: Single point of failure

Feedback: Single points of failure are indeed a challenge, but this option is not directly related to data
storage and management.

*C: Network latency

Feedback: Correct! Network latency can affect data retrieval and synchronization in distributed systems.

D: High costs of physical storage

Feedback: While cost is a consideration, it's not directly related to the management challenges within
distributed systems.

*E: Data redundancy

Feedback: Correct! Managing data redundancy to prevent data loss is a challenge in distributed systems.

Question 118 - checkbox, shuffle, partial credit, easy difficulty

Question category: Module: Module 2: Large Scale Data Storage

Which of the following are key features of Kafka's architecture?

*A: Kafka can handle a large number of simultaneous client connections.

Feedback: Correct! Kafka is designed to handle thousands of clients simultaneously.

B: Kafka stores data in a structured table format like SQL databases.

Feedback: This is incorrect. Kafka stores data in logs, not structured tables.

*C: Kafka supports horizontal scaling by adding more broker nodes.

Feedback: Correct! Kafka achieves scalability by adding more nodes to the cluster.

D: Kafka processes data using MapReduce.

Feedback: Incorrect. Kafka does not use MapReduce; it's a messaging system for real-time data
streaming.

*E: Kafka retains data for a configurable amount of time.

Feedback: Correct! Kafka allows configuring the retention period for stored data.

Question 119 - checkbox, shuffle, partial credit, easy difficulty

Question category: Module: Module 4: Graph Processing and Machine Learning

Select the advantages of learning cloud infrastructure and applications.

*A: Improved career opportunities

Feedback: Correct! Learning cloud infrastructure and applications can significantly improve your career
opportunities.

*B: Better understanding of cloud security

Feedback: Correct! Understanding cloud security is a key advantage of learning cloud infrastructure and
applications.

C: Ability to create mobile applications

Feedback: Incorrect. While you might learn related skills, creating mobile applications is not the primary
focus of this course.

D: Expertise in blockchain technology

Feedback: Incorrect. Blockchain technology is not covered in this course.

E: Increased networking skills

Feedback: Incorrect. While networking skills are beneficial, this course focuses on cloud infrastructure
and applications.

Question 120 - checkbox, shuffle, partial credit, easy difficulty

Question category: Module: Module 4: Graph Processing and Machine Learning

Which of the following are common use cases of big data machine learning?

*A: Predictive maintenance

Feedback: Correct! Predictive maintenance is a common use case of big data machine learning.

*B: Image recognition

Feedback: Correct! Image recognition is another common use case.

C: Document editing

Feedback: Incorrect. Document editing is not a common use case of big data machine learning. It is
typically handled by word processing software.

*D: Speech-to-text conversion

Feedback: Correct! Speech-to-text conversion is a common use case of big data machine learning.

E: Web browsing

Feedback: Incorrect. Web browsing itself is not a use case of big data machine learning, although big
data can be used to improve the experience.

Question 121 - checkbox, shuffle, partial credit, easy difficulty

Question category: Module: Module 4: Graph Processing and Machine Learning

Select the services that are typically part of a cloud infrastructure service model.

*A: Compute resources

Feedback: Correct. Compute resources are a fundamental part of cloud infrastructure services.

*B: Storage solutions

Feedback: Correct. Storage solutions are commonly provided in cloud infrastructure service models.

*C: Networking capabilities

Feedback: Correct. Networking capabilities are essential in cloud infrastructure services.

D: Personalized tech support

Feedback: Incorrect. Personalized tech support is generally an add-on service and not a core part of the
cloud infrastructure service model.

E: User interface customization

Feedback: Incorrect. User interface customization is usually a feature of SaaS applications, not core
cloud infrastructure services.

Question 122 - checkbox, shuffle, partial credit, easy difficulty

Question category: Module: Module 4: Graph Processing and Machine Learning

Select all frameworks that are used for graph processing.

*A: Pregel

Feedback: Correct! Pregel is a graph processing framework.

*B: Giraph

Feedback: Correct! Giraph is a graph processing framework.

*C: Spark GraphX

Feedback: Correct! Spark GraphX is a graph processing framework.

D: TensorFlow

Feedback: Incorrect. TensorFlow is primarily used for machine learning and deep learning.

E: PyTorch

Feedback: Incorrect. PyTorch is primarily used for machine learning and deep learning.

Question 123 - checkbox, shuffle, partial credit, easy difficulty

Question category: Module: Module 4: Graph Processing and Machine Learning

Select the key components typically included in a cloud infrastructure stack.

*A: Virtual Machines (VMs)

Feedback: Correct! VMs are a fundamental component of cloud infrastructure, providing scalable and
flexible compute resources.

B: Physical Servers

Feedback: Incorrect. While physical servers underlie cloud infrastructure, they are not considered part of
the cloud stack itself.

*C: Storage Solutions

Feedback: Correct! Storage solutions are essential for data management in cloud infrastructure.

*D: Networking Resources

Feedback: Correct! Networking resources are crucial for connectivity and communication within the
cloud.

E: Manual Configuration Tools

Feedback: Incorrect. Cloud infrastructure relies on automated tools rather than manual configuration.

F: Data Centers

Feedback: Incorrect. Data centers host cloud infrastructure but are not part of the stack itself.

Question 124 - checkbox, shuffle, partial credit, easy difficulty

Question category: Module: Module 4: Graph Processing and Machine Learning

Which of the following are practical applications of graph processing?

*A: Web-based computing

Feedback: Correct! Web-based computing is a significant practical application of graph processing.

*B: Transportation route optimization

Feedback: Correct! Transportation route optimization is a key practical application of graph processing.

C: Word processing

Feedback: Incorrect. Word processing is not typically a practical application of graph processing.
*D: Social networking

Feedback: Correct! Social networking heavily relies on graph processing.

E: Image editing

Feedback: Incorrect. Image editing is not a practical application of graph processing.

Question 125 - text match, easy difficulty

Question category: Module: Module 4: Graph Processing and Machine Learning

What is the term used to describe the process of automatically scaling cloud resources based on
demand? Please answer in all lowercase.

*A: autoscaling

Feedback: Correct. Autoscaling refers to the automatic adjustment of cloud resources based on demand.

*B: auto-scaling

Feedback: Correct. Auto-scaling is alternatively spelled with a hyphen but means the same.

C: scaling

Feedback: Incorrect. Scaling is a general term; the specific process is called autoscaling.

Default Feedback: Incorrect. Refer to the course material on how cloud resources can be automatically
adjusted based on demand.

Question 126 - numeric, easy difficulty

Question category: Module: Module 4: Graph Processing and Machine Learning

How many frameworks mentioned in this lesson are specifically used for graph processing?

*A: 3.0

Feedback: Correct! Pregel, Giraph, and Spark GraphX are the three frameworks mentioned.

Default Feedback: Incorrect. Try counting the frameworks mentioned in the lesson again.

Question 127 - text match, easy difficulty

Question category: Module: Module 4: Graph Processing and Machine Learning

What is the name of the algorithm used for clustering data points into groups in Mahout? Please answer
in all lowercase.

*A: kmeans

Feedback: Correct! The K-means algorithm is used for clustering data points into groups in Mahout.

*B: k-means

Feedback: Correct! The K-means algorithm is used for clustering data points into groups in Mahout.

Default Feedback: Incorrect. Refer to the material on clustering algorithms in Mahout.

Question 128 - text match, easy difficulty

Question category: Module: Module 4: Graph Processing and Machine Learning

What is one key benefit of cloud infrastructure that can be highlighted on a resume? Please answer in all
lowercase.

*A: scalability

Feedback: Correct! Scalability is a major benefit of cloud infrastructure.

Default Feedback: Incorrect. Please review the course material on the benefits of cloud infrastructure.

Question 129 - text match, easy difficulty

Question category: Module: Module 4: Graph Processing and Machine Learning

What framework is an open-source implementation of Pregel, initiated by Apache? Please answer in all
lowercase.

*A: giraph

Feedback: Correct! Giraph is the open-source implementation of Pregel.

B: graphx

Feedback: Incorrect. Please review the frameworks for graph processing in the lesson.

Default Feedback: Incorrect. Please review the frameworks for graph processing in the lesson.

Question 130 - numeric, easy difficulty

Question category: Module: Module 4: Graph Processing and Machine Learning

What is the typical availability percentage guaranteed by most cloud service providers for their
infrastructure services?

*A: 99.9

Feedback: Correct. Most cloud service providers guarantee a 99.9% availability for their infrastructure
services.

Default Feedback: Incorrect. Review the service level agreements (SLAs) commonly provided by cloud
service providers.

Question 131 - text match, easy difficulty

Question category: Module: Module 4: Graph Processing and Machine Learning

Name one framework specifically used for graph processing. Please answer in all lowercase.

*A: pregel

Feedback: Correct! Pregel is a framework specifically designed for graph processing.

*B: giraph

Feedback: Correct! Giraph is a framework specifically designed for graph processing.

*C: graphx

Feedback: Correct! Spark GraphX is another framework specifically used for graph processing.

Default Feedback: Incorrect. Please review the frameworks used for graph processing.

Question 132 - text match, easy difficulty

Question category: Module: Module 4: Graph Processing and Machine Learning

What is the master/worker model in Pregel responsible for? Please answer in all lowercase.

*A: distribution

Feedback: Correct! The master/worker model in Pregel is responsible for distributing work to workers.

B: allocation

Feedback: Incorrect. Please review the role of the master/worker model in Pregel.

C: management
Feedback: Incorrect. Please review the role of the master/worker model in Pregel.

Default Feedback: Incorrect. Please review the role of the master/worker model in Pregel.

Question 133 - text match, easy difficulty

Question category: Module: Module 4: Graph Processing and Machine Learning

What is the name of Apache Spark's graph processing module? Please answer in all lowercase.

*A: graphx

Feedback: Correct! GraphX is the graph processing module in Apache Spark.

B: graph

Feedback: Incorrect. Please review the name of Apache Spark's graph processing module.

C: graphspark

Feedback: Incorrect. Please review the name of Apache Spark's graph processing module.

Default Feedback: Incorrect. Please review the name of Apache Spark's graph processing module.

Question 134 - multiple choice, shuffle, easy difficulty

Question category: Module: Module 4: Graph Processing and Machine Learning

Which of the following components is essential for ensuring scalability in cloud infrastructure?

*A: Load balancers

Feedback: Correct! Load balancers help distribute traffic across multiple servers, ensuring scalability
and reliability.

B: Local storage devices

Feedback: Not quite. Local storage devices are not typically associated with scalability in cloud
environments.

C: Single server deployment

Feedback: This is incorrect. A single server deployment does not provide scalability.

D: Static IP addresses
Feedback: No, static IP addresses do not affect scalability.

Question 135 - multiple choice, shuffle, easy difficulty

Question category: Module: Module 4: Graph Processing and Machine Learning

Which algorithm is used for extracting frequent item sets efficiently in large datasets?

*A: Apriori Algorithm

Feedback: Correct! The Apriori Algorithm is used for finding frequent item sets by generating candidate
sets and checking their support.

B: K-means Clustering

Feedback: Not quite. K-means is used for clustering data points, not for extracting frequent item sets.

C: Naive Bayes Classifier

Feedback: Incorrect. The Naive Bayes Classifier is used for classification tasks, not for frequent item
extraction.

D: Support Vector Machine

Feedback: Incorrect. Support Vector Machine is used for classification and regression tasks, not for
extracting frequent item sets.

Question 136 - multiple choice, shuffle, easy difficulty

Question category: Module: Module 4: Graph Processing and Machine Learning

Which of the following is a key difference between machine learning and deep learning?

*A: Machine learning often requires manual feature extraction, while deep learning automates this
process.

Feedback: That's correct! Deep learning models automatically extract features through multiple layers,
whereas traditional machine learning models might require manual feature extraction.

B: Deep learning models are always faster to train than machine learning models.

Feedback: Not quite. Deep learning models can be computationally intensive and may not always train
faster than machine learning models.

C: Machine learning models are capable of understanding unstructured data without any preprocessing.
Feedback: Incorrect. Machine learning models typically require preprocessing of unstructured data,
unlike deep learning models.

D: Deep learning models do not require any labeled data for training purposes.

Feedback: This is not accurate. Deep learning models often require large amounts of labeled data to
perform effectively.

Question 137 - multiple choice, shuffle, easy difficulty

Question category: Module: Module 4: Graph Processing and Machine Learning

In cloud infrastructure, what is the primary role of a hypervisor?

*A: It manages virtual machines and allocates resources.

Feedback: Correct! The hypervisor is responsible for managing virtual machines and resource
allocation.

B: It provides an interface for application development.

Feedback: Incorrect. The primary role of a hypervisor is not related to application development
interfaces.

C: It ensures data encryption and security.

Feedback: Incorrect. While security is important, the primary role of a hypervisor is not focused on data
encryption.

D: It balances network load across servers.

Feedback: Incorrect. Load balancing is typically handled by other network management tools, not the
hypervisor.

Question 138 - multiple choice, shuffle, easy difficulty

Question category: Module: Module 4: Graph Processing and Machine Learning

Which of the following best describes a key benefit of including cloud infrastructure knowledge on your
resume?

*A: It demonstrates proficiency in modern technology stacks and enhances employability.

Feedback: Correct! Highlighting cloud skills shows your adaptability and technical prowess.

B: It guarantees a high-paying job in the tech industry.

Feedback: Not necessarily. While cloud skills increase employability, job guarantees depend on multiple
factors.

C: It replaces the need for other technical skills.

Feedback: Incorrect. While valuable, cloud skills should complement other technical proficiencies.

D: It allows you to bypass interviews and get hired immediately.

Feedback: Unlikely. While cloud skills are valuable, they do not eliminate the interview process.

Question 139 - multiple choice, shuffle, easy difficulty

Question category: Module: Module 4: Graph Processing and Machine Learning

Which of the following is a primary characteristic that differentiates machine learning from graph
processing?

*A: Machine learning focuses on predictive analytics while graph processing analyzes relationships.

Feedback: Correct! Machine learning is typically concerned with making predictions based on data,
whereas graph processing is more about understanding the connections and relationships between
entities.

B: Graph processing uses linear regression models predominantly.

Feedback: Not quite. Linear regression models are a staple in machine learning, not graph processing.

C: Machine learning algorithms require graph databases.

Feedback: This is incorrect. Machine learning algorithms require data but not necessarily in the form of
graph databases.

D: Graph processing and machine learning both fundamentally rely on deep learning techniques.

Feedback: Deep learning is a subset of machine learning and not a fundamental requirement for graph
processing.

Question 140 - multiple choice, shuffle, easy difficulty

Question category: Module: Module 4: Graph Processing and Machine Learning

How can completing the 'Cloud Infrastructure & Applications' course enhance your resume and future
career opportunities?

*A: By demonstrating expertise in cloud computing technologies.

Feedback: Correct! This course enhances your resume by showing your proficiency in cloud computing,
a highly sought-after skill.

B: By providing a certificate that guarantees job placement.

Feedback: Not quite. While a certificate is valuable, it does not guarantee job placement by itself.

C: By making you an eligible candidate for any IT role.

Feedback: Incorrect. This course focuses on cloud infrastructure and applications, not all IT roles.

D: By teaching coding skills applicable to all software development.

Feedback: Almost, but this course specifically focuses on cloud computing, not general software
development.

Question 141 - multiple choice, shuffle, medium

Question category: Module: Module 4: Graph Processing and Machine Learning

What is the main advantage of using the BSP model and Pregel for distributed graph processing
compared to MapReduce?

*A: It allows for iterative processing with better synchronization.

Feedback: Correct! The BSP model and Pregel enable iterative processing and efficient synchronization,
which are crucial for graph processing.

B: It reduces the computational complexity significantly.

Feedback: While complexity can be reduced, the main advantage is the ability to handle iterative
processing efficiently.

C: It uses less memory and CPU resources.

Feedback: Resource usage depends on the implementation. The key advantage lies in iterative
processing capabilities.

D: It simplifies the programming model for developers.

Feedback: Though it offers some simplification, the main benefit is in handling iterative processing
effectively.

Question 142 - multiple choice, shuffle, medium

Question category: Module: Module 4: Graph Processing and Machine Learning

What is one major advantage of using the Bulk Synchronous Parallel (BSP) model over the traditional
MapReduce model for distributed graph processing?

*A: BSP handles graph data with iterative processes more efficiently.

Feedback: Correct! BSP is designed to handle iterative processes, making it more suitable for graph data
processing compared to MapReduce.

B: MapReduce is designed for real-time processing of graph data.

Feedback: Incorrect. MapReduce is not optimized for iterative graph processing or real-time scenarios.

C: BSP uses a single-step processing model to optimize performance.

Feedback: Incorrect. BSP uses a multi-step processing model to facilitate iterative processing, unlike the
single-step model suggestion.

D: MapReduce has built-in support for graph-specific algorithms.

Feedback: Incorrect. MapReduce does not inherently support graph-specific algorithms; Pregel and BSP
are better suited for such tasks.

Question 143 - checkbox, shuffle, partial credit, easy difficulty

Question category: Module: Module 4: Graph Processing and Machine Learning

Which of the following are benefits of using Spark MLlib for machine learning?

*A: High scalability

Feedback: Correct! Spark MLlib is designed to handle large-scale data processing, making it highly
scalable.

*B: Integration with Hadoop

Feedback: Correct! Spark MLlib integrates well with Hadoop, providing a robust ecosystem for data
processing.

C: Automatic feature selection

Feedback: Not quite. While Spark MLlib offers many features, automatic feature selection is not one of
its primary benefits.

D: Support for deep learning frameworks

Feedback: Incorrect. While Spark MLlib is powerful, it is not specifically designed to support deep
learning frameworks directly.

E: Real-time data processing

Feedback: While Spark provides near real-time data processing, it is not the primary benefit of Spark
MLlib, which focuses more on batch processing for machine learning.

Question 144 - checkbox, shuffle, partial credit, hard

Question category: Module: Module 4: Graph Processing and Machine Learning

Which of the following are key components and phases involved in using Giraph for graph processing?

*A: Input splitting

Feedback: Correct! Input splitting is a key phase in Giraph processing.

*B: Vertex computation

Feedback: Correct! Vertex computation is a crucial component in Giraph's processing model.

*C: Output aggregation

Feedback: Correct! Output aggregation is an important phase in Giraph.

D: Map execution

Feedback: Map execution is not a phase specific to Giraph; it's more related to MapReduce.

E: Node synchronization

Feedback: Node synchronization is essential, but it's not uniquely categorized as a phase in Giraph.

Question 145 - checkbox, shuffle, partial credit, easy difficulty

Question category: Module: Module 4: Graph Processing and Machine Learning

In what ways can learning about cloud infrastructure and applications enhance your career? Select all
that apply.

*A: Improves problem-solving skills through cloud-based solutions.

Feedback: Correct! Cloud infrastructure knowledge can enhance your problem-solving skills.

B: Guarantees an immediate promotion at your current job.

Feedback: Incorrect. While beneficial, learning about cloud does not guarantee promotions.

*C: Expands understanding of scalable application deployment.

Feedback: Correct! Understanding cloud solutions can help you deploy applications more effectively.

D: Replaces need for understanding basic networking concepts.

Feedback: Incorrect. Cloud skills complement but do not replace foundational networking knowledge.

*E: Increases job market competitiveness in tech roles.

Feedback: Correct! Cloud expertise makes you a more competitive candidate in tech industries.

Question 146 - checkbox, shuffle, partial credit, medium

Question category: Module: Module 4: Graph Processing and Machine Learning

Select the key characteristics of cloud-native applications.

*A: Microservices architecture

Feedback: Correct! Cloud-native applications often utilize microservices architecture.

B: Monolithic design

Feedback: Incorrect. Cloud-native applications typically avoid monolithic designs in favor of

microservices.

*C: Containerization

Feedback: Correct! Containerization is a key characteristic of cloud-native applications.

D: Inflexibility to change

Feedback: Incorrect. Cloud-native applications are characterized by their flexibility and adaptability.

*E: Continuous integration and delivery

Feedback: Correct! Continuous integration and delivery are essential practices for cloud-native
applications.

Question 147 - checkbox, shuffle, partial credit, easy difficulty

Question category: Module: Module 4: Graph Processing and Machine Learning

Which platforms are used for graph processing? Select all that apply.

*A: Spark GraphX

Feedback: Correct! Spark GraphX is a powerful tool for graph processing.

*B: Pregel

Feedback: Correct! Pregel is designed specifically for large-scale graph processing.

*C: Giraph

Feedback: Correct! Giraph is an open-source platform for graph processing.

D: TensorFlow

Feedback: This is incorrect. TensorFlow is mainly used for machine learning and deep learning, not
graph processing.

E: PyTorch

Feedback: PyTorch is primarily used for machine learning, not graph processing.

Question 148 - checkbox, shuffle, partial credit, easy difficulty

Question category: Module: Module 4: Graph Processing and Machine Learning

Select the benefits of adding the 'Cloud Infrastructure & Applications' course to your resume.

*A: Enhances your understanding of cloud applications.

Feedback: Correct! Understanding cloud applications is a key takeaway from this course.

*B: Increases your chances of getting a cloud-related job.

Feedback: Correct! Cloud skills are in high demand, and this course can help you stand out.

C: Guarantees a higher salary in your next job.

Feedback: This might be possible, but the course itself does not guarantee a higher salary.

D: Provides networking opportunities with industry professionals.

Feedback: Incorrect. While learning platforms might offer networking events, this is not a direct benefit
of the course itself.
*E: Demonstrates commitment to professional development.

Feedback: Correct! Taking the course shows a proactive approach to learning and career growth.

Question 149 - checkbox, shuffle, partial credit, easy difficulty

Question category: Module: Module 4: Graph Processing and Machine Learning

Which of the following frameworks are used for graph processing?

*A: Pregel

Feedback: Correct! Pregel is one of the frameworks used for graph processing.

B: TensorFlow

Feedback: Not quite. TensorFlow is primarily used for deep learning, not graph processing.

*C: Giraph

Feedback: Well done! Giraph is another framework used for graph processing.

*D: Spark GraphX

Feedback: That's right! Spark GraphX is also used for graph processing.

E: PyTorch

Feedback: Incorrect. PyTorch is mainly used for machine learning and deep learning tasks, not
specifically for graph processing.

Question 150 - text match, easy difficulty

Question category: Module: Module 4: Graph Processing and Machine Learning

Name the algorithm that assumes independence between features for multi-class classification. Please
answer in all lowercase.

*A: naivebayes

Feedback: Correct! The Naive Bayes algorithm assumes independence between features and is used for
multi-class classification.

*B: naivebayesclassifier
Feedback: Correct! The Naive Bayes Classifier is a common implementation of the Naive Bayes
algorithm for classification tasks.

Default Feedback: Remember that the algorithm assumes feature independence and is commonly used
for classification.

Question 151 - text match, easy difficulty

Question category: Module: Module 4: Graph Processing and Machine Learning

What is the Google-developed system for large-scale graph processing? Please answer in all lowercase.

*A: pregel

Feedback: Correct! Pregel is Google's system for scalable graph processing.

B: pragel

Feedback: Close, but it seems like a typographical error. The correct term is a bit different.

Default Feedback: Review the graph processing systems discussed in the course material, focusing on
those developed by Google.

Question 152 - text match, easy difficulty

Question category: Module: Module 4: Graph Processing and Machine Learning

What machine learning library in Spark is used for clustering, classification, and regression? Please
answer in all lowercase.

*A: mllib

Feedback: Correct! MLlib is Spark’s machine learning library for clustering, classification, and
regression.

Default Feedback: Remember to review Spark's machine learning components to better understand what
MLlib offers.

Question 153 - text match, easy difficulty

Question category: Module: Module 4: Graph Processing and Machine Learning

Name one open-source platform specifically utilized for graph processing. Please answer in all
lowercase.

*A: giraph
Feedback: Correct! Giraph is an open-source platform used for graph processing.

*B: pregel

Feedback: Correct! Pregel is another platform used specifically for graph processing.

*C: sparkgraphx

Feedback: Correct! Spark GraphX is utilized for graph processing.

Default Feedback: Try again. Consider the platforms specifically designed for graph processing.

Question 154 - text match, easy difficulty

Question category: Module: Module 4: Graph Processing and Machine Learning

What is the term used to describe a model in Pregel that distributes tasks to multiple workers? Please
answer in all lowercase.

*A: master

Feedback: Correct! The master model is central to Pregel's task distribution system.

*B: master-worker

Feedback: Correct! The master-worker model is indeed used for task distribution in Pregel.

Default Feedback: Think about the hierarchical model used to manage and distribute tasks in Pregel.