Cloud Computing for Data Analysis - Midterm Exam - Practice Questions - Answer Key 


1 . In relation to the MapReduce environment , files are stored in a Distributed File System . Which of the following statements is incorrect in relation to the Distributed File System ?

C. The files are stored on a cluster of high performance data center machines
related to   slides 11 - 18  in  PowerPoint   MMS_01_Intro_MapReduce_Chapter01.pptx  


2 . Which of the following statements are example tasks for which MapReduce is suitable ?

i . ranking of Web pages by importance 
ii . searches in  friends   networks at social networking sites
iii . counting sales and grouping them by City or Region
iv . counting words in a text file

D . i . , ii . , iii . , and iv . 
related to   page 21  in  Chapter 2. from  Book 1. MiningOfMassiveDatasets  , and  VideoCase01 


3 . What is the purpose of the Reduce Tasks in MapReduce environment ?

B . Each Reducer produces some output , and the entire job is the union of what is produced by each Reducer 
related to  slide 21   in  PowerPoint    MMS_01_Intro_MapReduce_Chapter01.pptx   


4 . What is the main need to have Hadoop ?

A . Need to process huge datasets on large clusters of computers
related to    slide 4   in  PowerPoint    CloudToolsOverview.ppt    

5 . Which of the following statements are correct in Relation to Hadoop MapReduce , compared to traditional Relational Database Management System ( RDBMS ) ?

i . Hadoop MapReduce can handle much larger data size than traditional RDBMS
ii . Hadoop MapReduce provides both ‘ Interactive and Batch ‘ processing access
iii . traditional RDBMS provides updates of type  ‘ read and write many times ‘ , while Hadoop MapReduce provides ‘ write once read many times ‘
iv . traditional RDBMS has ‘ low ‘ integrity , while Hadoop MapReduce has ‘ high ‘ integrity 

B . i . and iii . 
related to   page 9   in   Chapter 1. from Book 2. HadoopTheDefinitiveGuide


6 . Which of the following statements is incorrect in relation to MapReduce ?

D . The number of Mappers , which are launched by Hadoop for a given job , is specified by the user 
related to    slides 24, 26, 27 , and 35   in  PowerPoint    CloudToolsOverview.ppt   


7 . What are the main goals of Hadoop Distributed File System ( HDFS ) ?

i . can be built out of commodity hardware 
ii . very large distributed file system suitable for applications with large datasets
iii . optimized for streaming data and iterative algorithm applications
iv . highly fault tolerant 

A . i . , ii . , and iv . 
related to    slide 8 ,   in  PowerPoint    CloudToolsOverview.ppt   , and 
slide 3  , in PowerPoint   HDFS.ppt  


8 . Hadoop Distributed File System ( HDFS ) stores large files across machines on a cluster . To do that , it uses the concept of a block . Which of the statements below is incorrect in relation to the HDFS block ?

C . The file metadata such as file name , owner of file , last modified date , are stored with the blocks
related to   pages 45 , 46    in Chapter 3. from Book 2. HadoopTheDefinitiveGuide


9 . What is the primary purpose of the NameNode in Hadoop Distributed File System ( HDFS ) ?

D . To manage the HDFS Namespace and Metadata , and keep an image of the entire file system namespace and map of blocks in memory
related to   pages 46 , 47    in Chapter 3. from Book 2. HadoopTheDefinitiveGuide   and 
slides 11 , 12   in  PowerPoint    CloudToolsOverview.ppt   , and 
slides 15 , 16   in PowerPoint    HDFS.ppt