Ccd-410 _ Cloudera Exams

(1)

12/23/13 Ccd-410 | Cloudera Exams

Cloudera Exams

Cloudera Certification Exams

Search Home Membership Home Aruba CheckPoint Cisco CIW Cloudera CompTIA CWNP EC-Council EMC EXIN HP IBM ISACA ISC2 Isilon Juniper Linux Microsoft Network Appliance Oracle PMI Sun VMware Zend EX AM I NFORMATI O N RSS feed for this section This category contains 60 posts

Exam CCD-410: Cloudera Certified Developer for Apache Hadoop which the reduce method of a given Reducer can be called?

When is the earliest point at which the reduce method of a given Reducer can be called? A.

As soon as at least one mapper has finished processing its input split. B.

As soon as a mapper has emitted at least one record. C.

Not until all mappers have finished processing all records.

D.

It depends on the InputFormat used for the job. Explanation:

In a MapReduce job reducers do not start executing the reduce method until the all

Map jobs have completed. Reducers start copying intermediate key-value pairs from the mappers as soon as they are available. The programmer defined reduce method is called only after all the mappers have finished.

Note: The reduce phase has 3 steps: shuffle, sort, reduce. Shuffle is where the data is collected by the reducer from each mapper. This can happen while mappers are generating data since it is only a data transfer. On the other hand, sort and reduce can only start once all the mappers are done. Why is starting the reducers early a good thing? Because it spreads out the data transfer from the mappers to the reducers over time, which is a good thing if your network is the bottleneck. Why is starting the reducers early a bad thing? Because they “hog up” reduce slots while only copying data. Another job that starts later that will actually use the reduce slots now can’t use them.

You can customize when the reducers startup by changing the default value of

mapred.reduce.slowstart.completed.maps in mapred-site.xml. A value of 1.00 will wait for all the mappers to finish before starting the reducers. A value of 0.0 will start the reducers right away. A value of 0.5 will start the reducers when half of the mappers are complete. You can also change mapred.reduce.slowstart.completed.maps on a job-by-job basis.

Typically, keep mapred.reduce.slowstart.completed.maps above 0.9 if the system ever has

(2)

multiple jobs running at once. This way the job doesn’t hog up reducers when they aren’t doing anything but copying data. If you only ever have one job running at a time, doing 0.1 would probably be appropriate.

Reference: 24 Interview Questions & Answers for Hadoop MapReduce developers, When is the reducers are started in a MapReduce job?

Which describes how a client reads a file from HDFS?

Which describes how a client reads a file from HDFS? A.

The client queries the NameNode for the block location(s). The NameNode returns the block location(s) to the client. The client reads the data directory off the DataNode(s).

B.

The client queries all DataNodes in parallel. The DataNode that contains the requested data responds directly to the client. The client reads the data directly off the DataNode. C.

The client contacts the NameNode for the block location(s). The NameNode then queries the DataNodes for block locations. The DataNodes respond to the NameNode, and the NameNode redirects the client to the DataNode that holds the requested data block(s). The client then reads the data directly off the DataNode.

D.

The client contacts the NameNode for the block location(s). The NameNode contacts the DataNode that holds the requested data block. Data is transferred from the DataNode to the NameNode, and then from the NameNode to the client.

Explanation:

The Client communication to HDFS happens using Hadoop HDFS API. Client

applications talk to the NameNode whenever they wish to locate a file, or when they want to add/copy/move/delete a file on HDFS. The NameNode responds the successful requests by returning a list of relevant DataNode servers where the data lives. Client applications can talk directly to a DataNode, once the NameNode has provided the location of the data.

Reference: 24 Interview Questions & Answers for Hadoop MapReduce developers, How the Client communicates with HDFS?

Which interface should your class implement?

You are developing a combiner that takes as input Text keys, IntWritable values, and emits Text keys, IntWritable values. Which interface should your class implement?

A.

Combiner <Text, IntWritable, Text, IntWritable> B.

Mapper <Text, IntWritable, Text, IntWritable> C.

Reducer <Text, Text, IntWritable, IntWritable> D.

Reducer <Text, IntWritable, Text, IntWritable>

E.

Combiner <Text, Text, IntWritable, IntWritable>

Indentify the utility that allows you to create and run MapReduce jobs with any executable or script as the mapper and/or the reducer?

A. Oozie B. Sqoop C. Flume

Posted by seenagape on October 26, 2013 No comments

(3)

12/23/13 Ccd-410 | Cloudera Exams D. Hadoop Streaming E. mapred Explanation:

Hadoop streaming is a utility that comes with the Hadoop distribution. The utility

allows you to create and run Map/Reduce jobs with any executable or script as the mapper and/or the reducer.

Reference: http://hadoop.apache.org/common/docs/r0.20.1/streaming.html (Hadoop Streaming, second sentence)

How are keys and values presented and passed to the reducers during a standard sort and shuffle phase of MapReduce?

A.

Keys are presented to reducer in sorted order; values for a given key are not sorted.

B.

Keys are presented to reducer in sorted order; values for a given key are sorted in ascending order.

C.

Keys are presented to a reducer in random order; values for a given key are not sorted. D.

Keys are presented to a reducer in random order; values for a given key are sorted in ascending order.

Explanation:

Reducer has 3 primary phases: 1. Shuffle

The Reducer copies the sorted output from each Mapper using HTTP across the network. 2. Sort

The framework merge sorts Reducer inputs by keys (since different Mappers may have output the same key).

The shuffle and sort phases occur simultaneously i.e. while outputs are being fetched they are merged.

SecondarySort

To achieve a secondary sort on the values returned by the value iterator, the application should extend the key with the secondary key and define a grouping comparator. The keys will be sorted using the entire key, but will be grouped using the grouping comparator to decide which keys and values are sent in the same call to reduce.

3. Reduce

In this phase the reduce(Object, Iterable, Context) method is called for each <key, (collection of values)> in the sorted inputs.

The output of the reduce task is typically written to a RecordWriter via TaskInputOutputContext.write(Object, Object).

The output of the Reducer is not re-sorted. Reference: org.apache.hadoop.mapreduce, Class Reducer<KEYIN,VALUEIN,KEYOUT,VALUEOUT>

Assuming default settings, which best describes the order of data provided to a reducer’s reduce method:

A.

The keys given to a reducer aren’t in a predictable order, but the values associated with those keys always are.

B.

Both the keys and values passed to a reducer always appear in sorted order.

(4)

C.

Neither keys nor values are in any predictable order. D.

The keys given to a reducer are in sorted order but the values associated with each key are in no predictable order

Explanation:

Reducer has 3 primary phases: 1. Shuffle

The Reducer copies the sorted output from each Mapper using HTTP across the network. 2. Sort

The framework merge sorts Reducer inputs by keys (since different Mappers may have output the same key).

The shuffle and sort phases occur simultaneously i.e. while outputs are being fetched they are merged.

SecondarySort

To achieve a secondary sort on the values returned by the value iterator, the application should extend the key with the secondary key and define a grouping comparator. The keys will be sorted using the entire key, but will be grouped using the grouping comparator to decide which keys and values are sent in the same call to reduce.

3. Reduce

In this phase the reduce(Object, Iterable, Context) method is called for each <key, (collection of values)> in the sorted inputs.

The output of the reduce task is typically written to a RecordWriter via TaskInputOutputContext.write(Object, Object).

The output of the Reducer is not re-sorted. Reference: org.apache.hadoop.mapreduce, Class Reducer<KEYIN,VALUEIN,KEYOUT,VALUEOUT>

Indentify the number of failed task attempts you can expect when you run the job with mapred.max.map.attempts set to 4:

You wrote a map function that throws a runtime exception when it encounters a control character in input data. The input supplied to your mapper contains twelve such characters totals, spread across five file splits. The first four file splits each have two control characters and the last split has four control characters.

Indentify the number of failed task attempts you can expect when you run the job with mapred.max.map.attempts set to 4:

A.

You will have forty-eight failed task attempts B.

You will have seventeen failed task attempts C.

You will have five failed task attempts D.

You will have twelve failed task attempts E.

You will have twenty failed task attempts

Explanation:

There will be four failed task attempts for each of the five file splits. Note:

which method in the Mapper you should use to implement code for reading the file and populating the associative array?

You want to populate an associative array in order to perform a map-side join. You’ve decided to put this information in a text file, place that file into the DistributedCache and read it in your Mapper before any records are processed.

(5)

12/23/13 Ccd-410 | Cloudera Exams Indentify which method in the Mapper you should use to implement code for reading the file and

populating the associative array? A. combine B. map C. init D. configure Explanation: See 3) below.

Here is an illustrative example on how to use the DistributedCache: // Setting up the cache for the application

1. Copy the requisite files to the FileSystem:

$ bin/hadoop fs -copyFromLocal lookup.dat /myapp/lookup.dat $ bin/hadoop fs -copyFromLocal map.zip /myapp/map.zip $ bin/hadoop fs -copyFromLocal mylib.jar /myapp/mylib.jar $ bin/hadoop fs -copyFromLocal mytar.tar /myapp/mytar.tar $ bin/hadoop fs -copyFromLocal mytgz.tgz /myapp/mytgz.tgz $ bin/hadoop fs -copyFromLocal mytargz.tar.gz /myapp/mytargz.tar.gz 2. Setup the application’s JobConf:

JobConf job = new JobConf();

DistributedCache.addCacheFile(new URI(“/myapp/lookup.dat#lookup.dat”), job);

DistributedCache.addCacheArchive(new URI(“/myapp/map.zip”, job); DistributedCache.addFileToClassPath(new Path(“/myapp/mylib.jar”), job); DistributedCache.addCacheArchive(new URI(“/myapp/mytar.tar”, job); DistributedCache.addCacheArchive(new URI(“/myapp/mytgz.tgz”, job); DistributedCache.addCacheArchive(new URI(“/myapp/mytargz.tar.gz”, job); 3. Use the cached files in the Mapper

or Reducer:

public static class MapClass extends MapReduceBase implements Mapper<K, V, K, V> {

private Path[] localArchives; private Path[] localFiles;

public void configure(JobConf job) { // Get the cached archives/files

localArchives = DistributedCache.getLocalCacheArchives(job); localFiles = DistributedCache.getLocalCacheFiles(job); }

public void map(K key, V value,

OutputCollector<K, V> output, Reporter reporter) throws IOException {

// Use data from the cached archives/files here // …

// …

output.collect(k, v); }

}

Reference: org.apache.hadoop.filecache , Class DistributedCache

which interface is most likely to reduce the amount of intermediate data transferred across the network?

You’ve written a MapReduce job that will process 500 million input records and generated 500 million key-value pairs. The data is not uniformly distributed. Your MapReduce job will create a significant amount of intermediate data that it needs to transfer between mappers and reduces which is a potential bottleneck. A custom implementation of which interface is most likely to reduce the amount of intermediate data transferred across the network?

A. Partitioner B.

OutputFormat

(6)

C. WritableComparable D. Writable E. InputFormat F. Combiner Explanation:

Combiners are used to increase the efficiency of a MapReduce program. They are

used to aggregate intermediate map output locally on individual mapper outputs. Combiners can help you reduce the amount of data that needs to be transferred across to the reducers. You can use your reducer code as a combiner if the operation performed is commutative and associative. Reference: 24 Interview Questions & Answers for Hadoop MapReduce developers, What are combiners? When should I use a combiner in my MapReduce Job?

Can you use MapReduce to perform a relational join on two large tables sharing a key?

Can you use MapReduce to perform a relational join on two large tables sharing a key? Assume that the two tables are formatted as comma-separated files in HDFS.

A.

Yes.

B.

Yes, but only if one of the tables fits into memory C.

Yes, so long as both tables fit into memory. D.

No, MapReduce cannot perform relational operations. E.

No, but it can be done with either Pig or Hive. Explanation:

Note:

* Join Algorithms in MapReduce A) Reduce-side join

B) Map-side join C) In-memory join

/ Striped Striped variant variant / Memcached variant * Which join to use?

/ In-memory join > map-side join > reduce-side join / Limitations of each?

In-memory join: memory

Map-side join: sort order and partitioning Reduce-side join: general purpose

Where is intermediate data written to after being emitted from the Mapper’s map method?

You have just executed a MapReduce job. Where is intermediate data written to after being emitted from the Mapper’s map method?

A.

Intermediate data in streamed across the network from Mapper to the Reduce and is never written to disk.

B.

Into in-memory buffers on the TaskTracker node running the Mapper that spill over and are written into HDFS.

C.

Into in-memory buffers that spill over to the local file system of the TaskTracker node running the Mapper.

D.

(7)

Into in-memory buffers that spill over to the local file system (outside HDFS) of the TaskTracker node running the Reducer

E.

Into in-memory buffers on the TaskTracker node running the Reducer that spill over and are written into HDFS.

Explanation:

The mapper output (intermediate data) is stored on the Local file system (NOT

HDFS) of each individual mapper nodes. This is typically a temporary directory location which can be setup in config by the hadoop administrator. The intermediate data is cleaned up after the Hadoop Job completes.

Reference: 24 Interview Questions & Answers for Hadoop MapReduce developers, Where is the Mapper Output (intermediate kay-value data) stored ?

How will you gather this data for your analysis?

You want to understand more about how users browse your public website, such as which pages they visit prior to placing an order. You have a farm of 200 web servers hosting your website. How will you gather this data for your analysis?

A.

Ingest the server web logs into HDFS using Flume. B.

Write a MapReduce job, with the web servers for mappers, and the Hadoop cluster nodes for reduces.

C.

Import all users’ clicks from your OLTP databases into Hadoop, using Sqoop. D.

Channel these clickstreams inot Hadoop using Hadoop Streaming. E.

Sample the weblogs from the web servers, copying them into Hadoop using curl. Explanation:

Hadoop MapReduce for Parsing Weblogs

Here are the steps for parsing a log file using Hadoop MapReduce: Load log files into the HDFS location using this Hadoop command: hadoop fs -put <local file path of weblogs> <hadoop HDFS location> The Opencsv2.3.jar framework is used for parsing log records.

Below is the Mapper program for parsing the log file from the HDFS location. public static class ParseMapper

extends Mapper<Object, Text, NullWritable,Text >{ private Text word = new Text();

public void map(Object key, Text value, Context context ) throws IOException, InterruptedException { CSVParser parse = new CSVParser(‘ ‘,’\”‘); String sp[]=parse.parseLine(value.toString()); int spSize=sp.length;

StringBuffer rec= new StringBuffer(); for(int i=0;i<spSize;i++){ rec.append(sp[i]); if(i!=(spSize-1)) rec.append(“,”); } word.set(rec.toString()); context.write(NullWritable.get(), word); } }

The command below is the Hadoop-based log parse execution. TheMapReduce program is attached in this article. You can add extra parsing methods in the class. Be sure to create a new JAR with any change and move it to the Hadoop distributed job tracker system.

hadoop jar <path of logparse jar> <hadoop HDFS logfile path> <output path of parsed log file> The output file is stored in the HDFS location, and the output file name starts with “part-”. which two issues?

(8)

MapReduce v2 (MRv2/YARN) is designed to address which two issues? A.

Single point of failure in the NameNode. B.

Resource pressure on the JobTracker.

C.

HDFS latency. D.

Ability to run frameworks other than MapReduce, such as MPI.

E.

Reduce complexity of the MapReduce APIs. F.

Standardize on a single MapReduce API. Explanation:

YARN (Yet Another Resource Negotiator), as an aspect of Hadoop, has two major kinds of benefits:

* (D) The ability to use programming frameworks other than MapReduce.

/ MPI (Message Passing Interface) was mentioned as a paradigmatic example of a MapReduce alternative

* Scalability, no matter what programming framework you use. Note:

* The fundamental idea of MRv2 is to split up the two major functionalities of the JobTracker, resource management and job scheduling/monitoring, into separate daemons. The idea is to have a global ResourceManager (RM) and per-application ApplicationMaster (AM). An application is either a single job in the classical sense of Map-Reduce jobs or a DAG of jobs.

* (B) The central goal of YARN is to clearly separate two things that are unfortunately smushed together in current Hadoop, specifically in (mainly) JobTracker:

/ Monitoring the status of the cluster with respect to which nodes have which resources available. Under YARN, this will be global.

/ Managing the parallelization execution of any specific job. Under YARN, this will be done separately for each job.

The current Hadoop MapReduce system is fairly scalable — Yahoo runs 5000 Hadoop jobs, truly concurrently, on a single cluster, for a total 1.5 – 2 millions jobs/cluster/month. Still, YARN will remove scalability bottlenecks

Reference: Apache Hadoop YARN – Concepts & Applications

which invocation correctly passes.mapred.job.name with a value of Example to Hadoop?

You need to run the same job many times with minor variations. Rather than hardcoding all job configuration options in your drive code, you’ve decided to have your Driver subclass org.apache.hadoop.conf.Configured and implement the org.apache.hadoop.util.Tool interface. Indentify which invocation correctly passes.mapred.job.name with a value of Example to Hadoop? A.

hadoop “mapred.job.name=Example” MyDriver input output B.

hadoop MyDriver mapred.job.name=Example input output C.

hadoop MyDrive –D mapred.job.name=Example input output

D.

hadoop setproperty mapred.job.name=Example MyDriver input output E.

hadoop setproperty (“mapred.job.name=Example”) MyDriver input output Explanation:

Configure the property using the -D key=value notation: -D mapred.job.name=’My Job’

You can list a whole bunch of options by calling the streaming jar with just the -info argument Reference: Python hadoop streaming : Setting a job name

(9)

Indentify what determines the data types used by the Mapper for a given job.

You are developing a MapReduce job for sales reporting. The mapper will process input keys representing the year (IntWritable) and input values representing product indentifies (Text). Indentify what determines the data types used by the Mapper for a given job.

A.

The key and value types specified in the JobConf.setMapInputKeyClass and JobConf.setMapInputValuesClass methods

B.

The data types specified in HADOOP_MAP_DATATYPES environment variable C.

The mapper-specification.xml file submitted with the job determine the mapper’s input key and value types.

D.

The InputFormat used by the job determines the mapper’s input key and value types.

Explanation:

The input types fed to the mapper are controlled by the InputFormat used. The

default input format, “TextInputFormat,” will load data in as (LongWritable, Text) pairs. The long value is the byte offset of the line in the file. The Text object holds the string contents of the line of the file.

Note: The data types emitted by the reducer are identified by setOutputKeyClass() andsetOutputValueClass(). The data types emitted by the reducer are identified by setOutputKeyClass() and setOutputValueClass().

By default, it is assumed that these are the output types of the mapper as well. If this is not the case, the methods setMapOutputKeyClass() and setMapOutputValueClass() methods of the JobConf class will override these.

Reference: Yahoo! Hadoop Tutorial, THE DRIVER METHOD

Identify the MapReduce v2 (MRv2 / YARN) daemon responsible for launching application containers and monitoring application resource usage?

A. ResourceManager B. NodeManager C. ApplicationMaster D. ApplicationMasterService E. TaskTracker F. JobTracker Explanation:

The fundamental idea of MRv2 (YARN) is to split up the two major functionalities of the JobTracker, resource management and job scheduling/monitoring, into separate daemons. The idea is to have a global ResourceManager (RM) and per-application ApplicationMaster (AM). An application is either a single job in the classical sense of Map-Reduce jobs or a DAG of jobs. Note: Let’s walk through an application execution sequence :

Which best describes how TextInputFormat processes input files and line breaks?

(10)

Which best describes how TextInputFormat processes input files and line breaks? A.

Input file splits may cross line breaks. A line that crosses file splits is read by the RecordReader of the split that contains the beginning of the broken line.

B.

Input file splits may cross line breaks. A line that crosses file splits is read by the RecordReaders of both splits containing the broken line.

C.

The input file is split exactly at the line breaks, so each RecordReader will read a series of complete lines.

D.

Input file splits may cross line breaks. A line that crosses file splits is ignored. E.

Input file splits may cross line breaks. A line that crosses file splits is read by the RecordReader of the split that contains the end of the broken line.

Explanation:

As the Map operation is parallelized the input file set is first split to several pieces

called FileSplits. If an individual file is so large that it will affect seek time it will be split to several Splits. The splitting does not know anything about the input file’s internal logical structure, for example line-oriented text files are split on arbitrary byte boundaries. Then a new map task is created per FileSplit.

When an individual map task starts it will open a new output writer per configured reduce task. It will then proceed to read its FileSplit using the RecordReader it gets from the specified InputFormat. InputFormat parses the input and generates key-value pairs. InputFormat must also handle records that may be split on the FileSplit boundary. For example TextInputFormat will read the last line of the FileSplit past the split boundary and, when reading other than the first FileSplit, TextInputFormat ignores the content up to the first newline.

Reference: How Map and Reduce operations are actually carried out For each input key-value pair, mappers can emit:

For each input key-value pair, mappers can emit: A.

As many intermediate key-value pairs as designed. There are no restrictions on the types of those key-value pairs (i.e., they can be heterogeneous).

B.

As many intermediate key-value pairs as designed, but they cannot be of the same type as the input key-value pair.

C.

One intermediate key-value pair, of a different type. D.

One intermediate key-value pair, but of the same type. E.

As many intermediate key-value pairs as designed, as long as all the keys have the same types and all the values have the same type.

Explanation:

Mapper maps input key/value pairs to a set of intermediate key/value pairs.

Maps are the individual tasks that transform input records into intermediate records. The transformed intermediate records do not need to be of the same type as the input records. A given input pair may map to zero or many output pairs.

Reference: Hadoop Map-Reduce Tutorial

How many keys will be passed to the Reducer’s reduce method?

You have the following key-value pairs as output from your Map task: (the, 1) (fox, 1) (faster, 1) (than, 1) (the, 1) (dog, 1)

(11)

12/23/13 Ccd-410 | Cloudera Exams How many keys will be passed to the Reducer’s reduce method?

A. Six B. Five C. Four D. Two E. One F. Three Explanation:

Only one key value pair will be passed from the two (the, 1) key value pairs. How will you obtain these user records?

You have user profile records in your OLPT database, that you want to join with web logs you have already ingested into the Hadoop file system. How will you obtain these user records? A.

HDFS command B.

Pig LOAD command

C.

Sqoop import D.

Hive LOAD DATA command E.

Ingest with Flume agents F.

Ingest with Hadoop Streaming Explanation:

Apache Hadoop and Pig provide excellent tools for extracting and analyzing data from very large Web logs.

We use Pig scripts for sifting through the data and to extract useful information from the Web logs. We load the log file into Pig using the LOAD command.

raw_logs = LOAD ‘apacheLog.log’ USING TextLoader AS (line:chararray); Note 1:

Data Flow and Components

* Content will be created by multiple Web servers and logged in local hard discs. This content will then be pushed to HDFS using FLUME framework. FLUME has agents running on Web servers; these are machines that collect data intermediately using collectors and finally push that data to HDFS.

* Pig Scripts are scheduled to run using a job scheduler (could be cron or any sophisticated batch job solution). These scripts actually analyze the logs on various dimensions and extract the results. Results from Pig are by default inserted into HDFS, but we can use storage

implementation for other repositories also such as HBase, MongoDB, etc. We have also tried the solution with HBase (please see the implementation section). Pig Scripts can either push this data to HDFS and then MR jobs will be required to read and push this data into HBase, or Pig scripts can push this data into HBase directly. In this article, we use scripts to push data onto HDFS, as we are showcasing the Pig framework applicability for log analysis at large scale.

* The database HBase will have the data processed by Pig scripts ready for reporting and further slicing and dicing.

* The data-access Web service is a REST-based service that eases the access and integrations with data clients. The client can be in any language to access REST-based API. These clients could be BI- or UI-based clients.

Note 2:

The Log Analysis Software Stack

(12)

* Hadoop is an open source framework that allows users to process very large data in parallel. It’s based on the framework that supports Google search engine. The Hadoop core is mainly divided into two modules:

1. HDFS is the Hadoop Distributed File System. It allows you to store large amounts of data using multiple commodity servers connected in a cluster.

2. Map-Reduce (MR) is a framework for parallel processing of large data sets. The default implementation is bonded with HDFS.

* The database can be a NoSQL database such as HBase. The advantage of a NoSQL database is that it provides scalability for the reporting module as well, as we can keep historical processed data for reporting purposes. HBase is an open source columnar DB or NoSQL DB, which uses HDFS. It can also use MR jobs to process data. It gives real-time, random read/write access to very large data sets — HBase can save very large tables having million of rows. It’s a distributed database and can also keep multiple versions of a single row.

* The Pig framework is an open source platform for analyzing large data sets and is implemented as a layered language over the Hadoop Map-Reduce framework. It is built to ease the work of developers who write code in the Map-Reduce format, since code in Map-Reduce format needs to be written in Java. In contrast, Pig enables users to write code in a scripting language.

* Flume is a distributed, reliable and available service for collecting, aggregating and moving a large amount of log data (src flume-wiki). It was built to push large logs into Hadoop-HDFS for further processing. It’s a data flow solution, where there is an originator and destination for each node and is divided into Agent and Collector tiers for collecting logs and pushing them to destination storage.

Reference: Hadoop and Pig for Large-Scale Web Log Analysis

What is the disadvantage of using multiple reducers with the default HashPartitioner and distributing your workload across you cluster?

A.

You will not be able to compress the intermediate data. B.

You will longer be able to take advantage of a Combiner. C.

By using multiple reducers with the default HashPartitioner, output files may not be in globally sorted order.

D.

There are no concerns with this approach. It is always advisable to use multiple reduces. Explanation:

Multiple reducers and total ordering

If your sort job runs with multiple reducers (either because mapreduce.job.reduces in mapredsite.xml has been set to a number larger than 1, or because you’ve used the -r option to specify

the number of reducers on the command-line), then by default Hadoop will use the HashPartitioner to distribute records across the reducers. Use of the HashPartitioner means that you can’t concatenate your output files to create a single sorted output file. To do this you’ll need total ordering,

Reference: Sorting text files with MapReduce

Which InputFormat should you use to complete the line: conf.setInputFormat (____.class) ; ?

Given a directory of files with the following structure: line number, tab character, string: Example:

1abialkjfjkaoasdfjksdlkjhqweroij 2kadfjhuwqounahagtnbvaswslmnbfgy 3kjfteiomndscxeqalkzhtopedkfsikj

You want to send each line as one record to your Mapper. Which InputFormat should you use to complete the line: conf.setInputFormat (____.class) ; ?

A. SequenceFileAsTextInputFormat B. SequenceFileInputFormat C. KeyValueFileInputFormat D.

(13)

12/23/13 Ccd-410 | Cloudera Exams BDBInputFormat

Explanation: Note:

The output format for your first MR job should be SequenceFileOutputFormat – this will store the Key/Values output from the reducer in a binary format, that can then be read back in, in your second MR job using SequenceFileInputFormat.

Reference: How to parse CustomWritable from text in Hadoop

http://stackoverflow.com/questions/9721754/how-to-parse-customwritable-from-text-in-hadoop (see answer 1 and then see the comment #1 for it)

Which is the best way to make this library available to your MapReducer job at runtime?

You need to perform statistical analysis in your MapReduce job and would like to call methods in the Apache Commons Math library, which is distributed as a 1.3 megabyte Java archive (JAR) file. Which is the best way to make this library available to your MapReducer job at runtime? A.

Have your system administrator copy the JAR to all nodes in the cluster and set its location in the HADOOP_CLASSPATH environment variable before you submit your job.

B.

Have your system administrator place the JAR file on a Web server accessible to all cluster nodes and then set the HTTP_JAR_URL environment variable to its location.

C.

When submitting the job on the command line, specify the –libjars option followed by the JAR file path.

D.

Package your code and the Apache Commands Math library into a zip file named JobJar.zip Explanation:

The usage of the jar command is like this, Usage: hadoop jar <jar> [mainClass] args…

If you want the commons-math3.jar to be available for all the tasks you can do any one of these

1. Copy the jar file in $HADOOP_HOME/lib dir or

2. Use the generic option -libjars. This is called:

The Hadoop framework provides a mechanism for coping with machine issues such as faulty configuration or impending hardware failure. MapReduce detects that one or a number of machines are performing poorly and starts more copies of a map or reduce task. All the tasks run simultaneously and the task finish first are used. This is called:

A. Combine B. IdentityMapper C. IdentityReducer D. Default Partitioner E. Speculative Execution Explanation:

Speculative execution: One problem with the Hadoop system is that by dividing the

tasks across many nodes, it is possible for a few slow nodes to rate-limit the rest of the program. For example if one node has a slow disk controller, then it may be reading its input at only 10% the speed of all the other nodes. So when 99 map tasks are already complete, the system is still waiting for the final map task to check in, which takes much longer than all the other nodes. By forcing tasks to run in isolation from one another, individual tasks do not know where their inputs come from. Tasks trust the Hadoop platform to just deliver the appropriate input. Therefore, the same input can be processed multiple times in parallel, to exploit differences in machine

(14)

capabilities. As most of the tasks in a job are coming to a close, the Hadoop platform will schedule redundant copies of the remaining tasks across several nodes which do not have other work to perform. This process is known as speculative execution. When tasks complete, they announce this fact to the JobTracker. Whichever copy of a task finishes first becomes the definitive copy. If other copies were executing speculatively, Hadoop tells the TaskTrackers to abandon the tasks and discard their outputs. The Reducers then receive their inputs from whichever Mapper completed successfully, first.

Reference: Apache Hadoop, Module 4: MapReduce Note:

* Hadoop uses “speculative execution.” The same task may be started on multiple boxes. The first one to finish wins, and the other copies are killed.

Failed tasks are tasks that error out.

* There are a few reasons Hadoop can kill tasks by his own decisions: a) Task does not report progress during timeout (default is 10 minutes)

b) FairScheduler or CapacityScheduler needs the slot for some other pool (FairScheduler) or queue (CapacityScheduler).

c) Speculative execution causes results of task not to be needed since it has completed on other place.

Reference: Difference failed tasks vs killed tasks For each intermediate key, each reducer task can emit:

For each intermediate key, each reducer task can emit: A.

As many final key-value pairs as desired. There are no restrictions on the types of those keyvalue pairs (i.e., they can be heterogeneous). B.

As many final key-value pairs as desired, but they must have the same type as the intermediate key-value pairs.

C.

As many final key-value pairs as desired, as long as all the keys have the same type and all the values have the same type.

D.

One final key-value pair per value associated with the key; no restrictions on the type. E.

One final key-value pair per key; no restrictions on the type.

Explanation:

Reducer reduces a set of intermediate values which share a key to a smaller set of values.

Reducing lets you aggregate values together. A reducer function receives an iterator of input values from an input list. It then combines these values together, returning a single output value. Reference: Hadoop Map-Reduce Tutorial; Yahoo! Hadoop Tutorial, Module 4: MapReduce What data does a Reducer reduce method process?

What data does a Reducer reduce method process? A.

All the data in a single input file. B.

All data produced by a single mapper. C.

All data for a given key, regardless of which mapper(s) produced it.

D.

All data for a given value, regardless of which mapper(s) produced it. Explanation:

Reducing lets you aggregate values together. A reducer function receives an iterator

of input values from an input list. It then combines these values together, returning a single output value.

All values with the same key are presented to a single reduce task. Reference: Yahoo! Hadoop Tutorial, Module 4: MapReduce

(15)

12/23/13 Ccd-410 | Cloudera Exams All keys used for intermediate output from mappers must:

All keys used for intermediate output from mappers must: A.

Implement a splittable compression algorithm. B. Be a subclass of FileInputFormat. C. Implement WritableComparable. D. Override isSplitable. E.

Implement a comparator for speedy sorting. Explanation:

The MapReduce framework operates exclusively on <key, value> pairs, that is, the

framework views the input to the job as a set of <key, value> pairs and produces a set of <key, value> pairs as the output of the job, conceivably of different types.

The key and value classes have to be serializable by the framework and hence need to implement the Writable interface. Additionally, the key classes have to implement the WritableComparable interface to facilitate sorting by the framework.

Reference: MapReduce Tutorial

What determines how the JobTracker assigns each map task to a TaskTracker?

On a cluster running MapReduce v1 (MRv1), a TaskTracker heartbeats into the JobTracker on your cluster, and alerts the JobTracker it has an open map task slot.

What determines how the JobTracker assigns each map task to a TaskTracker? A.

The amount of RAM installed on the TaskTracker node. B.

The amount of free disk space on the TaskTracker node. C.

The number and speed of CPU cores on the TaskTracker node. D.

The average system load on the TaskTracker node over the past fifteen (15) minutes. E.

The location of the InsputSplit to be processed in relation to the location of the node.

Explanation:

The TaskTrackers send out heartbeat messages to the JobTracker, usually every few minutes, to reassure the JobTracker that it is still alive. These message also inform the JobTracker of the number of available slots, so the JobTracker can stay up to date with where in the cluster work can be delegated. When the JobTracker tries to find somewhere to schedule a task within the MapReduce operations, it first looks for an empty slot on the same server that hosts the DataNode containing the data, and if not, it looks for an empty slot on a machine in the same rack.

Reference: 24 Interview Questions & Answers for Hadoop MapReduce developers, How JobTracker schedules a task?

which best defines a SequenceFile?

Indentify which best defines a SequenceFile? A.

A SequenceFile contains a binary encoding of an arbitrary number of homogeneous Writable objects

B.

A SequenceFile contains a binary encoding of an arbitrary number of heterogeneous Writable objects

(16)

C.

A SequenceFile contains a binary encoding of an arbitrary number of WritableComparable objects, in sorted order.

D.

A SequenceFile contains a binary encoding of an arbitrary number key-value pairs. Each key must be the same type. Each value must be the same type.

Explanation:

SequenceFile is a flat file consisting of binary key/value pairs. There are 3 different SequenceFile formats:

Uncompressed key/value records.

Record compressed key/value records – only ‘values’ are compressed here.

Block compressed key/value records – both keys and values are collected in ‘blocks’ separately and compressed. The size of the ‘block’ is configurable.

Reference: http://wiki.apache.org/hadoop/SequenceFile

which best describes the file access rules in HDFS if the file has a single block that is stored on data nodes A, B and C?

A client application creates an HDFS file named foo.txt with a replication factor of 3. Identify which best describes the file access rules in HDFS if the file has a single block that is stored on data nodes A, B and C?

A.

The file will be marked as corrupted if data node B fails during the creation of the file. B.

Each data node locks the local file to prohibit concurrent readers and writers of the file. C.

Each data node stores a copy of the file in the local file system with the same name as the HDFS file.

D.

The file can be accessed if at least one of the data nodes storing the file is available.

Explanation:

HDFS keeps three copies of a block on three different datanodes to protect against

true data corruption. HDFS also tries to distribute these three replicas on more than one rack to protect against data availability issues. The fact that HDFS actively monitors any failed datanode(s) and upon failure detection immediately schedules re-replication of blocks (if needed) implies that three copies of data on three different nodes is sufficient to avoid corrupted files. Note:

HDFS is designed to reliably store very large files across machines in a large cluster. It stores each file as a sequence of blocks; all blocks in a file except the last block are the same size. The blocks of a file are replicated for fault tolerance. The block size and replication factor are configurable per file. An application can specify the number of replicas of a file. The replication factor can be specified at file creation time and can be changed later. Files in HDFS are write-once and have strictly one writer at any time. The NameNode makes all decisions regarding replication of blocks. HDFS uses rack-aware replica placement policy. In default configuration there are total 3 copies of a datablock on HDFS, 2 copies are stored on datanodes on same rack and 3rd copy on a different rack.

Reference: 24 Interview Questions & Answers for Hadoop MapReduce developers , How the HDFS Blocks are replicated?

how many blocks the input file occupies?

In a MapReduce job, you want each of your input files processed by a single map task. How do you configure a MapReduce job so that a single map task processes each input file regardless of how many blocks the input file occupies?

A.

Increase the parameter that controls minimum split size in the job configuration. B.

Write a custom MapRunner that iterates over all key-value pairs in the entire file. C.

Set the number of mappers equal to the number of input files you want to process. D.

Write a custom FileInputFormat and override the method isSplitable to always return false.

(17)

12/23/13 Ccd-410 | Cloudera Exams Explanation:

FileInputFormat is the base class for all file-based InputFormats. This provides a

generic implementation of getSplits(JobContext). Subclasses of FileInputFormat can also override the isSplitable(JobContext, Path) method to ensure input-files are not split-up and are processed as a whole by Mappers.

Reference: org.apache.hadoop.mapreduce.lib.input, Class FileInputFormat<K,V> Which process describes the lifecycle of a Mapper?

Which process describes the lifecycle of a Mapper? A.

The JobTracker calls the TaskTracker’s configure () method, then its map () method and finally its close () method.

B.

The TaskTracker spawns a new Mapper to process all records in a single input split. C.

The TaskTracker spawns a new Mapper to process each key-value pair.

D.

The JobTracker spawns a new Mapper to process all records in a single file. Explanation:

For each map instance that runs, the TaskTracker creates a new instance of your mapper.

Note:

* The Mapper is responsible for processing Key/Value pairs obtained from the InputFormat. The mapper may perform a number of Extraction and Transformation functions on the Key/Value pair before ultimately outputting none, one or many Key/Value pairs of the same, or different Key/Value type.

* With the new Hadoop API, mappers extend the org.apache.hadoop.mapreduce.Mapper class. This class defines an ‘Identity’ map function by default – every input Key/Value pair obtained from the InputFormat is written out.

Examining the run() method, we can see the lifecycle of the mapper: /**

* Expert users can override this method for more complete control over the * execution of the Mapper.

* @param context * @throws IOException */

public void run(Context context) throws IOException, InterruptedException { setup(context);

while (context.nextKeyValue()) {

map(context.getCurrentKey(), context.getCurrentValue(), context); }

cleanup(context); }

setup(Context) – Perform any setup for the mapper. The default implementation is a no-op method. map(Key, Value, Context) – Perform a map operation in the given Key / Value pair. The default implementation calls Context.write(Key, Value)

cleanup(Context) – Perform any cleanup for the mapper. The default implementation is a no-op method.

Reference: Hadoop/MapReduce/Mapper

which best describes when the reduce method is first called in a MapReduce job?

Determine which best describes when the reduce method is first called in a MapReduce job? A.

Reducers start copying intermediate key-value pairs from each Mapper as soon as it has completed. The programmer can configure in the job what percentage of the intermediate data should arrive before the reduce method begins.

B.

Reducers start copying intermediate key-value pairs from each Mapper as soon as it has completed. The reduce method is called only after all intermediate data has been copied and sorted.

(18)

C.

Reduce methods and map methods all start at the beginning of a job, in order to provide optimal performance for map-only or reduce-only jobs.

D.

Reducers start copying intermediate key-value pairs from each Mapper as soon as it has completed. The reduce method is called as soon as the intermediate key-value pairs start to arrive.

Explanation:

* In a MapReduce job reducers do not start executing the reduce method until the all

Map jobs have completed. Reducers start copying intermediate key-value pairs from the mappers as soon as they are available. The programmer defined reduce method is called only after all the mappers have finished.

* Reducers start copying intermediate key-value pairs from the mappers as soon as they are available. The progress calculation also takes in account the processing of data transfer which is done by reduce process, therefore the reduce progress starts showing up as soon as any intermediate key-value pair for a mapper is available to be transferred to reducer. Though the reducer progress is updated still the programmer defined reduce method is called only after all the mappers have finished.

Reference: 24 Interview Questions & Answers for Hadoop MapReduce developers , When is the reducers are started in a MapReduce job?

How many times will the Reducer’s reduce method be invoked?

You have written a Mapper which invokes the following five calls to the OutputColletor.collect method:

output.collect (new Text (“Apple”), new Text (“Red”) ) ; output.collect (new Text (“Banana”), new Text (“Yellow”) ) ; output.collect (new Text (“Apple”), new Text (“Yellow”) ) ; output.collect (new Text (“Cherry”), new Text (“Red”) ) ; output.collect (new Text (“Apple”), new Text (“Green”) ) ; How many times will the Reducer’s reduce method be invoked? A. 6 B. 3 C. 1 D. 0 E. 5 Explanation:

reduce() gets called once for each [key, (list of values)] pair. To explain, let’s say you called:

out.collect(new Text(“Car”),new Text(“Subaru”); out.collect(new Text(“Car”),new Text(“Honda”); out.collect(new Text(“Car”),new Text(“Ford”); out.collect(new Text(“Truck”),new Text(“Dodge”); out.collect(new Text(“Truck”),new Text(“Chevy”); Then reduce() would be called twice with the pairs reduce(Car, <Subaru, Honda, Ford>)

reduce(Truck, <Dodge, Chevy>) Reference: Mapper output.collect()? What is the best way to accomplish this?

To process input key-value pairs, your mapper needs to lead a 512 MB data file in memory. What is the best way to accomplish this?

A.

Serialize the data file, insert in it the JobConf object, and read the data into memory in the configure method of the mapper.

(19)

12/23/13 Ccd-410 | Cloudera Exams B.

Place the data file in the DistributedCache and read the data into memory in the map method of the mapper.

C.

Place the data file in the DataCache and read the data into memory in the configure method of the mapper.

D.

Place the data file in the DistributedCache and read the data into memory in the configure method of the mapper.

Explanation:

Hadoop has a distributed cache mechanism to make available file locally that may be needed by Map/Reduce jobs

Use Case

Lets understand our Use Case a bit more in details so that we can follow-up the code snippets. We have a Key-Value file that we need to use in our Map jobs. For simplicity, lets say we need to replace all keywords that we encounter during parsing, with some other value.

So what we need is

A key-values files (Lets use a Properties files) The Mapper code that uses the code Write the Mapper code that uses it view sourceprint?

01.

public class DistributedCacheMapper extends Mapper<LongWritable, Text, Text, Text> { 02. 03. Properties cache; 04. 05. @Override 06.

protected void setup(Context context) throws IOException, InterruptedException { 07.

super.setup(context); 08.

Path[] localCacheFiles = DistributedCache.getLocalCacheFiles(context.getConfiguration()); 09.

10.

if(localCacheFiles != null) { 11.

// expecting only single file here 12.

for (int i = 0; i < localCacheFiles.length; i++) { 13.

Path localCacheFile = localCacheFiles[i]; 14.

cache = new Properties(); 15. cache.load(new FileReader(localCacheFile.toString())); 16. } 17. } else { 18.

// do your error handling here 19. } 20. 21. } 22. 23. @Override 24.

public void map(LongWritable key, Text value, Context context) throws IOException, InterruptedException {

25.

// use the cache here 26.

(20)

27.

// do some action or replace with something else 28. } 29. 30. } Note:

* Distribute application-specific large, read-only files efficiently.

DistributedCache is a facility provided by the Map-Reduce framework to cache files (text, archives, jars etc.) needed by applications.

Applications specify the files, via urls (hdfs:// or http://) to be cached via the JobConf. The DistributedCache assumes that the files specified via hdfs:// urls are already present on the FileSystem at the path specified by the url.

Reference: Using Hadoop Distributed Cache

Which statement best describes the ordering of these values?

In a MapReduce job, the reducer receives all values associated with same key. Which statement best describes the ordering of these values?

A.

The values are in sorted order. B.

The values are arbitrarily ordered, and the ordering may vary from run to run of the same MapReduce job.

C.

The values are arbitrary ordered, but multiple runs of the same MapReduce job will always have the same ordering.

D.

Since the values come from mapper outputs, the reducers will receive contiguous sections of sorted values.

Explanation: Note:

* Input to the Reducer is the sorted output of the mappers.

* The framework calls the application’s Reduce function once for each unique key in the sorted order.

* Example:

For the given sample input the first map emits: < Hello, 1>

< World, 1> < Bye, 1> < World, 1> The second map emits: < Hello, 1> < Hadoop, 1> < Goodbye, 1> < Hadoop, 1>

which two resources should you expect to be bottlenecks?

You need to create a job that does frequency analysis on input data. You will do this by writing a Mapper that uses TextInputFormat and splits each value (a line of text from an input file) into individual characters. For each one of these characters, you will emit the character as a key and an InputWritable as the value. As this will produce proportionally more intermediate data than input data, which two resources should you expect to be bottlenecks?

A.

Processor and network I/O B.

Disk I/O and network I/O

C.

Processor and RAM D.

Processor and disk I/O

(21)

12/23/13 Ccd-410 | Cloudera Exams Will you be able to reuse your existing Reduces as your combiner in this case and why or why not?

You want to count the number of occurrences for each unique word in the supplied input data. You’ve decided to implement this by having your mapper tokenize each word and emit a literal value 1, and then have your reducer increment a counter for each literal 1 it receives. After successful implementing this, it occurs to you that you could optimize this by specifying a combiner. Will you be able to reuse your existing Reduces as your combiner in this case and why or why not?

A.

Yes, because the sum operation is both associative and commutative and the input and output types to the reduce method match.

B.

No, because the sum operation in the reducer is incompatible with the operation of a Combiner. C.

No, because the Reducer and Combiner are separate interfaces. D.

No, because the Combiner is incompatible with a mapper which doesn’t use the same data type for both the key and value.

E.

Yes, because Java is a polymorphic object-oriented language and thus reducer code can be reused as a combiner.

Explanation:

used to aggregate intermediate map output locally on individual mapper outputs. Combiners can help you reduce the amount of data that needs to be transferred across to the reducers. You can use your reducer code as a combiner if the operation performed is commutative and associative. The execution of combiner is not guaranteed, Hadoop may or may not execute a combiner. Also, if required it may execute it more then 1 times. Therefore your MapReduce jobs should not depend on the combiners execution.

Reference: 24 Interview Questions & Answers for Hadoop MapReduce developers, What are combiners? When should I use a combiner in my MapReduce Job?

Identify the Hadoop daemon on which the Hadoop framework will look for an available slot schedule a MapReduce operation.

Your client application submits a MapReduce job to your Hadoop cluster. Identify the Hadoop daemon on which the Hadoop framework will look for an available slot schedule a MapReduce operation. A. TaskTracker B. NameNode C. DataNode D. JobTracker E. Secondary NameNode Explanation:

JobTracker is the daemon service for submitting and tracking MapReduce jobs in

Hadoop. There is only One Job Tracker process run on any hadoop cluster. Job Tracker runs on its own JVM process. In a typical production cluster its run on a separate machine. Each slave node is configured with job tracker node location. The JobTracker is single point of failure for the Hadoop MapReduce service. If it goes down, all running jobs are halted. JobTracker in Hadoop performs following actions(from Hadoop Wiki:)

Client applications submit jobs to the Job tracker.

The JobTracker talks to the NameNode to determine the location of the data The JobTracker locates TaskTracker nodes with available slots at or near the data The JobTracker submits the work to the chosen TaskTracker nodes.

The TaskTracker nodes are monitored. If they do not submit heartbeat signals often enough, they are deemed to have failed and the work is scheduled on a different TaskTracker.

(22)

A TaskTracker will notify the JobTracker when a task fails. The JobTracker decides what to do then: it may resubmit the job elsewhere, it may mark that specific record as something to avoid, and it may may even blacklist the TaskTracker as unreliable.

When the work is completed, the JobTracker updates its status. Client applications can poll the JobTracker for information.

Reference: 24 Interview Questions & Answers for Hadoop MapReduce developers, What is a JobTracker in Hadoop? How many instances of JobTracker run on a Hadoop Cluster?

Which project gives you a distributed, Scalable, data store that allows you random, realtime read/write access to hundreds of terabytes of data?

A. HBase B. Hue C. Pig D. Hive E. Oozie F. Flume G. Sqoop Explanation:

Use Apache HBase when you need random, realtime read/write access to your Big Data.

Note: This project’s goal is the hosting of very large tables — billions of rows X millions of columns – atop clusters of commodity hardware. Apache HBase is an open-source, distributed, versioned, column-oriented store modeled after Google’s Bigtable: A Distributed Storage System for Structured Data by Chang et al. Just as Bigtable leverages the distributed data storage provided by the Google File System, Apache HBase provides Bigtable-like capabilities on top of Hadoop and HDFS.

Features

Linear and modular scalability. Strictly consistent reads and writes. Automatic and configurable sharding of tables Automatic failover support between RegionServers.

Convenient base classes for backing Hadoop MapReduce jobs with Apache HBase tables. Easy to use Java API for client access.

Block cache and Bloom Filters for real-time queries. Query predicate push down via server side Filters

Thrift gateway and a REST-ful Web service that supports XML, Protobuf, and binary data encoding options

Extensible jruby-based (JIRB) shell

Support for exporting metrics via the Hadoop metrics subsystem to files or Ganglia; or via JMX Reference: http://hbase.apache.org/ (when would I use HBase? First sentence)

what would another user see when trying to access this life?

You use the hadoop fs –put command to write a 300 MB file using and HDFS block size of 64 MB. Just after this command has finished writing 200 MB of this file, what would another user see when trying to access this life?

A.

They would see Hadoop throw an ConcurrentFileAccessException when they try to access this file.

B.

They would see the current state of the file, up to the last bit written by the command. C.

(23)

12/23/13 Ccd-410 | Cloudera Exams They would see the current of the file through the last completed block.

D.

They would see no content until the whole file written and closed.

Explanation: Note: * put

Usage: hadoop fs -put <localsrc> … <dst>

Copy single src, or multiple srcs from local file system to the destination filesystem. Also reads input from stdin and writes to destination filesystem.

Identify the tool best suited to import a portion of a relational database every day as files into HDFS, and generate Java classes to interact with that imported data?

A. Oozie B. Flume C. Pig D. Hue E. Hive F. Sqoop G. fuse-dfs Explanation:

Sqoop (“SQL-to-Hadoop”) is a straightforward command-line tool with the following capabilities: Imports individual tables or entire databases to files in HDFS

Generates Java classes to allow you to interact with your imported data

Provides the ability to import from SQL databases straight into your Hive data warehouse Note:

Data Movement Between Hadoop and Relational Databases

Data can be moved between Hadoop and a relational database as a bulk data transfer, or relational tables can be accessed from within a MapReduce map function.

Note:

* Cloudera’s Distribution for Hadoop provides a bulk data transfer tool (i.e., Sqoop) that imports individual tables or entire databases into HDFS files. The tool also generates Java classes that support interaction with the imported data. Sqoop supports all relational databases over JDBC, and Quest Software provides a connector (i.e., OraOop) that has been optimized for access to data residing in Oracle databases.

Reference: http://log.medcl.net/item/2011/08/hadoop-and-mapreduce-big-data-analytics-gartner/ (Data Movement between hadoop and relational databases, second paragraph)

How many files will be processed by the FileInputFormat.setInputPaths () command when it’s given a path object representing this directory?

You have a directory named jobdata in HDFS that contains four files: _first.txt, second.txt, .third.txt and #data.txt. How many files will be processed by the FileInputFormat.setInputPaths () command when it’s given a path object representing this directory?

A.

Four, all files will be processed B.

Three, the pound sign is an invalid character for HDFS file names C.

Two, file names with a leading period or underscore are ignored

(24)

D.

None, the directory cannot be named jobdata E.

One, no special characters can prefix the name of an input file Explanation:

Files starting with ‘_’ are considered ‘hidden’ like unix files starting with ‘.’. # characters are allowed in HDFS file names.

Determine the difference between setting the number of reduces to one and settings the number of reducers to zero.

You write MapReduce job to process 100 files in HDFS. Your MapReduce algorithm uses TextInputFormat: the mapper applies a regular expression over input values and emits key-values pairs with the key consisting of the matching text, and the value containing the filename and byte offset. Determine the difference between setting the number of reduces to one and settings the number of reducers to zero.

A.

There is no difference in output between the two settings. B.

With zero reducers, no reducer runs and the job throws an exception. With one reducer, instances of matching patterns are stored in a single file on HDFS.

C.

With zero reducers, all instances of matching patterns are gathered together in one file on HDFS. With one reducer, instances of matching patterns are stored in multiple files on HDFS. D.

With zero reducers, instances of matching patterns are stored in multiple files on HDFS. With one reducer, all instances of matching patterns are gathered together in one file on HDFS.

Explanation:

* It is legal to set the number of reduce-tasks to zero if no reduction is desired.

In this case the outputs of the map-tasks go directly to the FileSystem, into the output path set by setOutputPath(Path). The framework does not sort the map-outputs before writing them out to the FileSystem.

* Often, you may want to process input data using a map function only. To do this, simply set mapreduce.job.reduces to zero. The MapReduce framework will not create any reducer tasks. Rather, the outputs of the mapper tasks will be the final output of the job.

Note: Reduce

In this phase the reduce(WritableComparable, Iterator, OutputCollector, Reporter) method is called for each <key, (list of values)> pair in the grouped inputs.

The output of the reduce task is typically written to the FileSystem via OutputCollector.collect(WritableComparable, Writable).

Applications can use the Reporter to report progress, set application-level status messages and update Counters, or just indicate that they are alive.

The output of the Reducer is not sorted. A combiner reduces:

A combiner reduces: A.

The number of values across different keys in the iterator supplied to a single reduce method call.

B.

The amount of intermediate data that must be transferred between the mapper and reducer.

C.

The number of input files a mapper must process. D.

The number of output files a reducer must produce. Explanation:

used to aggregate intermediate map output locally on individual mapper outputs. Combiners can help you reduce the amount of data that needs to be transferred across to the reducers. You can

(25)

12/23/13 Ccd-410 | Cloudera Exams use your reducer code as a combiner if the operation performed is commutative and associative. The execution of combiner is not guaranteed, Hadoop may or may not execute a combiner. Also, if required it may execute it more then 1 times. Therefore your MapReduce jobs should not depend on the combiners execution.

Reference: 24 Interview Questions & Answers for Hadoop MapReduce developers, What are combiners? When should I use a combiner in my MapReduce Job?

how many map task attempts will there be?

In a MapReduce job with 500 map tasks, how many map task attempts will there be? A.

It depends on the number of reduces in the job. B. Between 500 and 1000. C. At most 500. D. At least 500. E. Exactly 500. Explanation:

From Cloudera Training Course:

Task attempt is a particular instance of an attempt to execute a task – There will be at least as many task attempts as there are tasks – If a task attempt fails, another will be started by the JobTracker

– Speculative execution can also result in more task attempts than completed tasks which major functions of the JobTracker into separate daemons?

MapReduce v2 (MRv2/YARN) splits which major functions of the JobTracker into separate daemons? Select two.

A.

Heath states checks (heartbeats) B.

Resource management

C.

Job scheduling/monitoring D.

Job coordination between the ResourceManager and NodeManager

E.

Launching tasks F.

Managing file system metadata G.

MapReduce metric reporting H.

Managing tasks Explanation:

The fundamental idea of MRv2 is to split up the two major functionalities of the

JobTracker, resource management and job scheduling/monitoring, into separate daemons. The idea is to have a global ResourceManager (RM) and per-application ApplicationMaster (AM). An application is either a single job in the classical sense of Map-Reduce jobs or a DAG of jobs. Note:

The central goal of YARN is to clearly separate two things that are unfortunately smushed together in current Hadoop, specifically in (mainly) JobTracker:

/ Monitoring the status of the cluster with respect to which nodes have which resources available. Under YARN, this will be global.

(26)

/ Managing the parallelization execution of any specific job. Under YARN, this will be done separately for each job.

What types of algorithms are difficult to express in MapReduce v1 (MRv1)?

What types of algorithms are difficult to express in MapReduce v1 (MRv1)? A.

Algorithms that require applying the same mathematical function to large numbers of individual binary records.

B.

Relational operations on large amounts of structured and semi-structured data. C.

Algorithms that require global, sharing states.

D.

Large-scale graph algorithms that require one-step link traversal. E.

Text analysis algorithms on large collections of unstructured text (e.g, Web crawls). Explanation:

See 3) below.

Limitations of Mapreduce – where not to use Mapreduce

While very powerful and applicable to a wide variety of problems, MapReduce is not the answer to every problem. Here are some problems I found where MapReudce is not suited and some papers that address the limitations of MapReuce.

1. Computation depends on previously computed values

If the computation of a value depends on previously computed values, then MapReduce cannot be used. One good example is the Fibonacci series where each value is summation of the previous two values. i.e., f(k+2) = f(k+1) + f(k). Also, if the data set is small enough to be computed on a single machine, then it is better to do it as a single reduce(map(data)) operation rather than going through the entire map reduce process.

2. Full-text indexing or ad hoc searching

The index generated in the Map step is one dimensional, and the Reduce step must not generate a large amount of data or there will be a serious performance degradation. For example, CouchDB’s MapReduce may not be a good fit for full-text indexing or ad hoc searching. This is a problem better suited for a tool such as Lucene.

3. Algorithms depend on shared global state

Solutions to many interesting problems in text processing do not require global synchronization. As a result, they can be expressed naturally in MapReduce, since map and reduce tasks run independently and in isolation. However, there are many examples of algorithms that depend crucially on the existence of shared global state during processing, making them difficult to implement in MapReduce (since the single opportunity for global synchronization in MapReduce is the barrier between the map and reduce phases of processing)

Reference: Limitations of Mapreduce – where not to use Mapreduce What does calling the next () method return?

In the reducer, the MapReduce API provides you with an iterator over Writable values. What does calling the next () method return?

A.

It returns a reference to a different Writable object time. B.

It returns a reference to a Writable object from an object pool. C.

It returns a reference to the same Writable object each time, but populated with different data.

D.

It returns a reference to a Writable object. The API leaves unspecified whether this is a reused object or a new object.

E.

It returns a reference to the same Writable object if the next value is the same as the previous value, or a new Writable object otherwise.