Hadoop Imp Commands

Hadoop Hackers Groups hadoophackers@googlegroups.
com
HADOOP COMMANDS
To Start Hadoop (Present Working Directory should be Hadoop Folder)
bin/start-all.sh
If you are inside BIN folder of Hadoop Folder

./start-all.sh
To Start Hadoop (Present Working Directory should be Hadoop Folder)

bin/stop-all.sh
./stop-all.sh
To start HDFS Daemon process start-dfs.sh -> To Stop: stop-dfs.sh

To start MR Daemon process start-mapred .sh -> To Stop: stop-mapred.sh
Hadoop allows to start/stop each and every daemons by below commands.
Daemon Process Command

Name Node hadoop-daemon.sh start namenode
Data Node hadoop-daemon.sh start datanode
Job Tracker hadoop-daemon.sh start jobtracker
Task Tracker hadoop-daemon.sh start tasktracker
Secondary Name Node hadoop-daemon.sh start secondarynamenode
To stop the respective daemons, use stop instead of start. Ex
hadoop-daemon.sh stop secondarynamenode
Note:
HDFS Daemons NN, SNN, JT
MR Daemons DN, TT
To Delete Name Node

bin/hadoop namenode -format
Note: This command should be EXECUTED only when the Hadoop is getting installed. It is
advised not to use this command once data are inserted into HDFS
To create Secure Key

ssh-keygen -t rsa
To Create Directory
hadoop dfs -mkdir /NewDirectory/Dir1
To Copy File
From Local System to HDFS
End of Page 1/21

Hadoop Hackers Groups hadoophackers@googlegroups.com
bin/hadoop dfs -copyFromLocal /home/administrator/R3.jar /ClientJavaF
From HDFS to Local

bin/hadoop dfs -copyToLocal /ClientJavaFolder/HadoopAddFile.txt
/home/administrator/Desktop/Hadoop/NewFileCopied.txt
From HDFS to HDFS

bin/hadoop dfs -cp /ClientJavaFolder/HadoopAddFile.txt /myNewDir/hello
Note: Directory name should not have SPACE or SPECIAL characters [ ( , ) , & , : , < , > ]
To Move : File will be copied and file from copied path will be deleted.
From Local System to HDFS

bin/hadoop dfs -moveFromLocal
/home/administrator/Desktop/Hadoop/NewFileCopied.txt
/NewFile/MovedFile.txt
From HDFS to HDFS

/hadoop dfs -mv /NewFile/MovedFile.txt /a.txt
To list the files available

bin/hadoop dfs -ls /
Note: between ls and / - space must be provided.
To Delete from HDFS
To Delete File inside Dir

bin/hadoop dfs -rm /ClientJavaFolder/NewRenameFile.txt
To Delete Directory
bin/hadoop dfs -rm /ClientJavaFolder
To Delete Directory which start with same name and ends with different character.
bin/hadoop dfs -rm /ClientJavaFolder/File*
Ex: If dir ClientJavaFolder have 3 files. If you want to delete all the files which starts with File
then you need to use above command. i.e. * will be used.
File1.txt
File2.txt
HelloFile.txt
To Rename the File name

bin/hadoop dfs -mv /ClientJavaFolder/HadoopAddFile.txt
/ClientJavaFolder/RenamedFile.txt
To View Directory View

bin/hadoop dfs -du /
End of Page 2/21

Ex:
bin/hadoop dfs -du /
Found 2 items
238925722 hdfs://localhost:50000/ClientJavaFolder
119462861 hdfs://localhost:50000/NewFile
To set Access permission

To set permission for File
bin/hadoop dfs -chmod 771 /ClientJavaFolder/NewRenameFile.txt
To set permission for Directory
bin/hadoop dfs -chmod 771 /ClientJavaFolder/
Note: Only Directories will be given execute permission and for all files READ WRITE can be
given.
To run JAR file

Use below CMD if the JAR is NOT EXPORTED along with class name
bin/hadoop jar /home/administrator/Print.jar
org.samples.mapreduce.training.patientrxMR /NewJavaFolder/NewRenameFile.txt
/hello.txt
Format:
bin/hadoop jar [jarpath] [classPath] [inputpath] [outputpath]
Use below CMD if the JAR is EXPORTED along with class name
bin/hadoop jar /home/administrator/Print.jar
/NewJavaFolder/NewRenameFile.txt /hello.txt
To run Streaming-API or other program files apart from JAVA.

hadoop jar /home/administrator/hadoop-1.2.1/contrib/streaming/hadoop-*streaming*.jar
\-file /home/administrator/hadoop-1.2.1/streaming_api/mapper.py -mapper
/home/administrator/hadoop-1.2.1/streaming_api/mapper.py \-file
/home/administrator/hadoop-1.2.1/streaming_api/reducer.py -reducer
/home/administrator/hadoop-1.2.1/streaming_api/reducer.py \-input /my30Mb_file
-output /my30Mb_file_output_1
Format:
bin/hadoop jar [streamingjar path]
\-file
[mapper type file format ]
-mapper
[mapper type file]
\-file
[reducer type file format ]
- reducer
[reducer type file]
-input
[input file path]
-output
End of Page 3/21

[output file path]
Note: Spaces are not allowed in above command for all BACKWARD SLASH
Ex: \ -file is wrong
\-file is corrent
The whole comment should be ONE LINE command
Generic Parserer Used via PRE-DEFINED Command (Using LOCAL HOST method)
hadoop jar /home/administrator/genericparserfiles.jar -files

hdfs://localhost:50000/GenericParsers/input2 /file /GenericParsers/OutputwithFiles2
Generic Parserer Used via HADOOP File Command
hadoop jar /home/administrator/genericparser.jar /hello.xt /GenericParsers/input2

/GenericParsers/Output3
Hadoop Safe Mode Command

hadoop dfsadmin -safemode leave
End of Page 4/21

HIVE
Without Partition;
create table patient(pid INT,pname STRING,drug STRING,gender STRING,tot_amt INT)
row format delimited fields terminated by ',' stored as textfile;
WithPartition;
create table country(cid INT, name STRING,cntry STRING, joindate STRING) partitioned
by(country STRING, jdate STRING) row format delimited fields terminated by ',' stored as
textfile;
To load data in HIVE;
From HDFS:
load data inpath '/my30Mb_file' into table patient;
From Local
load data local inpath '/home/administrator/Desktop/US' into table patient;
Partitioned by:
load data local inpath '/home/administrator/Desktop/US' into table country partition
(country='US',jdate='01-05-2010');
Note: All partition variables must be used when trying to load data. Ex country and jdate are
partition variables. If we try to use country along, it will not work. We need to use all varaibles.
To write output in HDFS

To HDFS:
insert overwrite directory '/hive/output' select * from patient;
To Local
insert overwrite local directory '/home/administrator/h' select * from patient;
(Use above query when you need only one column to be stored. For more columns
storage, see below note section)
Note: Format problem will be available if we try to store in local drive. i.e. data will be stored
without spaces or any delimiters. We can use below query to put our own delimiter.
insert overwrite local directory '/home/administrator/Desktop/Show' row format

delimited fields terminated by ',' select * from country;
This delimiter can be any thing irrespective to default delimiter which was given in CREATE
statement.
End of Page 5/21

Partitions in HIVE
From HDFS:
load data inpath '/home/administrator/Desktop/US' into table country partition
From Local
load data local inpath '/home/administrator/Desktop/US' into table country partition
To Run User Defined Functions
To Add Jar:
add jar file:///home/administrator/hive.jar
To Add Jar:
create temporary function myfn as 'org.samples.hive.training.hiveUDAF';
To start HWI service

bin/hive --service hwi
Note: Hive will allow only one instance of HWI and in mulitple terminals if you try to execute this
query. You will end up with error.
To Start Hive Server

bin/hive --service hiveserver
Buckets:
create table bucket (cid INT, cname STRING) CLUSTERED BY(cid) 3 buckets row format
delimited fields terminated by ',' stored as textfile;
Note: When you try to load the same data again to same directory, it will create separate instance.
Whereas in normal method without bucketing, it will replace with the loaded data.
To load data in Bucket:

insert overwrite table bkt_table_name select * from table_name;
End of Page 6/21

HBASE
To Start Hbase (Present Working Directory should be Hbase Folder)
bin/start-hbase.sh
If you are inside BIN folder of HBase Folder

./start-hbase .sh
To Start Hbase (Present Working Directory should be Hbase Folder)

bin/stop-hbase .sh
If you are inside BIN folder of Hbase Folder

./stop-hbase .sh
To enter HBASE Prompt

bin/hbase shell
To show tables use below comment

list
To create table
create 'mercury','Personal','Medical'
Note: Here QUERY

mercury is TABLE name
Personal & Medical Family
Below are some other methods to create table

1. create 'table1', {NAME => 'f1'}
1. NAME Attirbute contributes as Family.
2. This follows KEY VALUE Pair method.
3. This attribute name must be in UPPER CASE.
2. create 't1', 'f1', {SPLITS => ['10', '20', '30', '40']}
To insert in table
put 'mercury','001','Personal:Name','Bala'
Note:
1. By Default, we will have 3 versions of data. It is configurable via code itself. Below
example shows how versions will be changed.
create 'sam1', {NAME => 'f1'},{NAME =>'f2',VERSIONS => 5}
2. If the table is disabled, put statement will not work. To find whether the table is
enabled or disabled. Use below command
is_disabled 'table_name'
To retrive values
Using GET Method
get 'mercury','001'
End of Page 7/21

To get specific column

Ex: get 'mercury','001' ,'Personal'
Ex: get 'mercury','001' ,'Personal:Age'
To get all columns available in table1

scan 'mercury'
Note: This will display latest version. To display all versions of data, use below command.
To get versions of tables

scan 'mercury',{VERSIONS =>3}
To delete values
delete 'dummy','002','f1:name' (specific family wise)
delete 'dummy','002' (specific row key wise)
To drop table
drop 'mercury'
Note: To delete we need to disble the table. Use disable 'mercury' to disable the table.
Enable/Disable/Drop/Exist
Description Command & Desc

Cmd Description
Enable Table enable 'table_name' Only after enabling the table, it is
allowed to write data.
To Enable more than one table enable_all 'tab.*'
To check whether the table is is_enabled 'table_name' Returns True if it is enabled else
disabled False.
Disable Table disable 'table_name' Data Insertion NOT POSSIBLE
To Disable more than one table disable_all 'tab.*'
To check whether the table is is_disabled 'table_name' Returns True if it is disabled else
disabled False.
To check whether table exist in exists 'table_name' Returns True if table exist else False.
system
To drop all tables drop_all 'tab.*' Note: All the table must be disabled.
End of Page 8/21

PIG
To Start PIG Service
To load data from HDFS
bin/pig
To load data from Local System
bin/pig -x local
Note:
1. If the mode is HDFS,
1. The output will be available in /tmp directory
2. If the mode is LOCAL,
1. Output will be available in /tmp folder. (click COMPUTER link in left pane, to see
this folder.
To Create & Load Data without Schema;

Login as
bin/pig -x local
To Load Data
A = load '/data10' using PigStorage(',');
A = load '/data10' using PigStorage(',') as (pid:int, pname:chararray, drug:chararray,
gender:chararray, amt:int);
To Filter Data
B = filter A by $2 =='avil';
To Create & Load Data with Schema(via HDFS);
Login as
bin/pig
To Load Data
C= load '/data10' using PigStorage(',') as (pid:int, pname:chararray,
drug:chararray, gender:chararray, tot_amt:int);
To Filter Data
D = filter C by drug=='avil'; (we use column values rather than $ value.)
X = filter C by (f1==8) OR (NOT (f2+f3 > f1));
To Fetch Specific Column Data

A = load '/data10' using PigStorage(',') as (pid:int, pname:chararray, drug:chararray,
gender:chararray, amt:int);
B = foreach A generate pid,drug; (using column values.)
C = foreach A generate $0,$3; (using $ variable.)
To Display Data
dump A
To Describe Bag
End of Page 9/21

illustrate A
To take Procedural Plan Data

explain A (procedureal Plan will be displayed.)
To Display Data
illustrate A
To Store Data in HDFS

Store F into '/Pig_Results' using PigStorage(',');
To Use Group Fn
G = GROUP F by drug;
sm = foreach G generate group,SUM(C.tot_amt) as S;
To Use CoGroup Fn
A = LOAD '/data10' using PigStorage(',');
B = LOAD '/drug' using PigStorage(',');
C = COGROUP A by $2, B by $1;
To use NOT NULL Condition:

B = filter A by $2 is not null;
To Use Distinct
A = LOAD '/tuple2' using PigStorage(',') as (f1:int, f2:int, f3:int);
B = GROUP A by f1;
C = foreach B generate group,SUM(A.f2) as S;
dump C
Split Fn
A = LOAD '/patientHR-10' using PigStorage(',') ;
SPLIT A into x if $0%2==0, z if $0==1;
dump x
dump y
Note: This can be used to do mulitple operations at a single query. Like Male records in one
Variable and Female records in another varaible.
To Use MACRO
DEFINE myMacro1(fTable,fCol,fValue) returns returnVariable
{ $returnVariable = FILTER $fTable BY $fCol == '$fValue';};
To Run the Macro

D = myMacro1(C,'drug','metacin');
To Use Joins
A = load '/data10' using PigStorage(',');
B = load '/drug' using PigStorage(',');
C = join A by $2, B by $1;
End of Page 10/21

To Use Inner/outer/left/right joins use below query

C = join A by $2 left outer, B by $1;
(left outer/right outer//full outer)
Project-Range Expressions (..) is used.

B = foreach A generate $0..$4;
Note: .. refers from $0th Column to $4th Column, the collumn
To Sort Data
B = foreach A generate $0..$4;
C = order B by $4;
C = order B by $4..$6;
C = order B by $4..$6,$7;
To run jar File

register /home/administrator/myFile.jar
B = foreach C generate com.Upper(pname);
Note: A, B, C & D are temporary variables to handle the table.
To Execute jar file instead of REGISTER method

A = LOAD '/WordcountInput.txt';
B = MAPREDUCE 'wordcount.jar' STORE A INTO 'inputDir' LOAD 'outputDir'
AS (word:chararray, count: int) `org.myorg.WordCo
Tuples/Bag/Maps
Tuple:
A = LOAD '/tuple' using PigStorage(' ') AS (t1:tuple(t1a:int,
t1b:int,t1c:int),t2:tuple(t2a:int,t2b:int,t2c:int));
(Refer PIG.txt, search above query for more explaination)
To Use Tuple
B = FOREACH A GENERATE t1.t1a,t2.t2a;
Bag:
(Refer PIG.txt, search above query for more explaination)
To Use Tuple
B = FOREACH A GENERATE t1.t1a,t2.t2a;
End of Page 11/21

Conventions of TUPLE/BAG/MAP:
(1,2,3) is tuple
{1,2} is bag
{ (1,3),(3,4) } is tuple inside bag
(123,10,{ (1,3),(3,4) } )is bag inside tuple
[key#value] is MAP
Ex: [ 'name' # 'John', 'ext' # 5555 ]
Bag:
End of Page 12/21

SQOOP
To check SQOOP & DB Connectivitiy
bin/sqoop
eval --connect jdbc:mysql://localhost/test
-username root
-password root
--query
"select * from patient";
To List Databases
bin/sqoop list-databases --connect jdbc:mysql://localhost/information_schema -username
root -password root
To Move Data from RDBMS to HDFS (Default Path: /user/username/table_name)

bin/sqoop
import
--connect jdbc:mysql://localhost/test
-username root
-password root
--table
patient
-m 1
-m Mapper
To Move Data from RDBMS to Customized HDFS Directory

bin/sqoop
import
-username root
-password root
--table patient
--split-by pid
-m 1 OR num-mappers (-num-mappers can be used instead of -m)
--target-dir /Sqoop (to store in HDFS dir)
--where pid < 10 (to use where clause condiitons)
Commands Descriptions
--query This command is used to accept the sql query. If you use this command, --table
should not be used.
--columns <col> If you need to fetch specific columns of table. Use this command.
Ex: --columns col1,col2,col3
--num-mappers It is as same as like -m option which we will provide to introduce how many
mappers will be running.
--warehouse-dir As same like target-dir, But the difference is it will create a directory of
End of Page 13/21

TABLE name and will be storing all details inside the newly created dir.
Ex: Consider PatientMR is the table which we are going to store in HDFS
Command Input Path HDFS Directory

--target-dir /My_Target /My_Target
--warehouse-dir /My_Target /My_Target/PatientMR
--where condition To give condiitons to filter the required records.

--append This command is used to append the records in same directory. If you do NOT
use this command, you cannot write the data in same path. If you use this
command, the data will written in same path with different file name and NOT
OVERWRITTING the data.
If the destination directory already exists in HDFS, Sqoop will

refuse to import and overwrite that directorys contents. If you use the
--append argument, Sqoop will import
data to a temporary directory and then rename the files into the normal target
directory which will not disturb existing directory files.
-z or -compress To compress the data and store it in HDFS as .gz file format.
--fields-terminated-by To store in HDFS with field needs to be delimiter.
[character] --fields-terminated-by \\t - Tab Space
--fields-terminated-by \\b Back Space
\\n newline\
You can use any character you need.

This will take only first character even if you give as word.
Ex: --fields-terminated-by XYZ
--hadoop-home To Override Hadoop Home
Generic Options Like Hadoop Commands, you can use generic options in SQOOP.
-D <property=value> use value for given property
-fs <local|namenode:port> specify a namenode

-jt <local|jobtracker:port> specify a job tracker
-files <comma separated list of files> specify comma separated files to be
copied to the map reduce cluster
-libjars <comma separated list of jars> specify comma separated jar files to
include in the classpath.
-archives <comma separated list of specify comma separated archives to
archives> be unarchived on the compute
End of Page 14/21

machines.
-conf <configuration file> specify an application configuration
file
To Import All Tables

bin/sqoop
import-all-tables
-username root
-password root
-m 1;
To Import Data to HIVE

bin/sqoop
import
-username root
-password root
--table patient
--hive-table patient_hive
--create-hive-table
--hive-import -m 1;
Import Commands
--input-escaped-by <char> Sets the input escape character
--input-fields-terminated-by <char> Sets the input field separator
--input-lines-terminated-by <char> Sets the input end-of-line character
--input-optionally-enclosed-by <char> Sets a field enclosing character
To Import Data to HBASE

bin/sqoop import
-username root
-password root
--table patient
--hbase-table patientsqp
--column-family MyFamily
--hbase-row-key pid
--hbase-create-table -m 1
To Export Data from HDFS to SQL

bin/sqoop
export
End of Page 15/21

--connect jdbc:mysql://localhost/test -username root -password root

--table patient1
--export-dir /Sqoop/AsTextFile/patient/part-m-00000
-m 1;
Export & Other Commands
--update-mode allowinsert By adding this command, allows user to insert the new records which
are not available in
--update-key <column name> This is used to update the table based on column name that is given.
--verbose Print more information while working
To Create Scoop JOB

bin/sqoop job
--create myjob
-- import
connect jdbc:mysql://localhost/test
-username root -password root
--table patient
-m 1 --target-dir /Sqoop/MRJobLastValue
--last-value pid
bin/sqoop job --list To display all the Sqoop JOB that are created.
bin/sqoop job exec <jobname> To execute the job created.
SQOOP Main Commands
codegen Generate code to interact with database records

create-hive-table Import a table definition into Hive
eval Evaluate a SQL statement and display the results
export Export an HDFS directory to a database table
help List available commands
import Import a table from a database to HDFS
import-all-tables Import tables from a database to HDFS
job Work with saved jobs
list-databases List available databases on a server
list-tables List available tables in a database
merge Merge results of incremental imports
metastore Run a standalone Sqoop metastore
version Display version information
End of Page 16/21

FLUME
To RUN Flume
bin/flume-ng
agent
--conf-file netcat_flume.conf
--name a1
-Dflume.root.logger=INFO,console
--conf-file To specify the file name we need to run in FLUME.

--name -- To specific the agent name. There are more than one agent can be available in conf file.
Thats why we specify Agent Name.
Sources
*-mandatory attributes
type, bind, port, interceptor
Attributes Description
type*
What type of transfer method the flume uses is mentioned. Ex:
a1.sources.r1.type = netcat (Terminal based)
a1.sources.r1.type = exec (Log File based)
To mention Log File, below code will be used. Also need to give batchsize.
a1.sources.r1.command = tail - F /home/administration/data.log
a1.sources.r1.batchSize = 2
(Batchsize is max no of lines to read and send to the channel at a time)
a1.sources.r1.type = spooldir (Directory based)
bind*
IP Address of Source machine or if the local machine is used. Then below method is
followed.
a1.sources.r1.bind = localhost
port* Port number is mentioned here.
Interceptors
Interceptors will be intermediater between Source and Channels.We need to specify type
and header of the interceptor.
a1.sources.r1.interceptors = i1
a1.sources.r1.interceptors.i1.type = host
a1.sources.r1.interceptors.i1.hostheader = hostname
TBD
End of Page 17/21

Multi Node Cluster Setup

Steps for Multi Node Cluster
1. Choose number of nodes in the cluster and assign the NN, SNN, JT, DN, TT respectively.
2. Generate SSH Key in NN. Copy the ssh key to all DN.
1. ssh-keygen -t rsa
2. Go to .ssh directory
3. cat id_rsa.pub >> authorized_keys
4. Now use below command to copy into DN.
1. Ssh-copy-id -i /home/adminstrator/.ssh/id_rsa.pub systemname/username
3. In all nodes, specify NN's IP Address in Master file and DN's IP Address in Slave file which is avialable in
below path.
1. Go to directory of HDFS. Say hadoop-1.2.1/conf. File called masters and slaves will be avaible.
2. Masters file will have default value localhost. We need to give IP address of NN.
3. Masters file will have default value localhost. We need to give IP address of DN.
4. In DN, open core-site.xml which is inside conf folder. Configure its NN value as NN Node IP address and
port number to access NN.
5. In DN, open mapred-site.xmlwhich is inside conf folder. Configure its JT value as JP Node IP address and
port number to access NN.
6. In DN, open hdfs-site.xml which is inside conf folder.
1. Set name dir, data dir, block size, replication factor.
7. Start Cluster and check whether the cluster is up.
Hadoop Admin Access Details

When do we need to go towards HADOOP?
If we have High Scalability & High Computation, then we need to implement HADOOP.
Steps for Admin
1. Design the Cluster

1. How many DN nodes needs to be used. Replication factors and others.
2. DesignHardware configuration
1. This should be in ratio of 1:2:4 i.e. 1 TB data : 2 Quad Core Processor : 4 GB RAM
2. RAM Size should be LARGE.
3. RAID Hard Drives will be used.
3. OS Selection
1. Ubuntu, Fedora Core, CentOS, RHEL and Solaris.
4. Java Version needs to selected.
1. Hadoop accepts all versions of JAVA expect Version 7 and Version 6 - 1.6.0u18.
5. Configure Hadoop details and Check the health of the cluster. Run below commands for checking.
1. bin/hadoop fsck /
2. bin/hadoop fsck / -files -blocks -racks
End of Page 18/21

Adding a node in LIVE Cluster.
1. In hdfs-site.xml and mapred-site.xml below changes needs to be made.

1. Below configuration needs to be done. i.e. setting include file
hdfs-site.xml
<property>
<name>dfs.hosts.include </name>
<value>/home/hadoop/includes</value>
<final>true</final>
</property>
mapred-site.xml
<property>
<name>mapred.hosts.include </name>
<value>/home/hadoop/includes</value>
<final>true</final>
</property>
2. Need to Update Name Node:

1. bin/hadoop dfsadmin -refreshNodes in DN machine.
2. Now in Commisionning Link, we will get the DN related details.
3. Again run same command to finaize.
3. In slaves file, update the new node IP address.
4. Start DN and TT in newly added node.
1. hadoop-daemon.sh start tasktracker
2. hadoop-daemon.sh start datanode
5. Cross check UI.
6. Run Balancer using this command start-balancer.sh
7. Refer below image
1. Deleting a node in LIVE Cluster.

1. In hdfs-site.xml and mapred-site.xml below changes needs to be made.
End of Page 19/21

2. Below configuration needs to be done. i.e. setting exclude file

hdfs-site.xml
<property>
<name>dfs.hosts.exclude </name>
<value>/home/hadoop/excludes</value>
<final>true</final>
</property>
mapred-site.xml
<property>
<name>mapred.hosts. exclude </name>
<value>/home/hadoop/excludes</value>
<final>true</final>
</property>
3. Need to Update Name Node:
1. bin/hadoop dfsadmin -refreshNodes in DN machine.
4. In slaves file, update the new node IP address.
5. Start DN and TT in newly added node.
1. hadoop-daemon.sh stop tasktracker
2. hadoop-daemon.sh stop datanode
6. Cross check UI.
7. Run Balancer using this command start-balancer.sh
8. Refer below image
Port Numbers
End of Page 20/21

Daemon Process Web Port RPC Port

Namenode 50070 8020 (50000)
Secondarynamenode 50090 -
Data Node 50075 50010
Job Tracker 50030 8021(50001)
Task Tracker 50060 50020
HMaster 60000 60010
HRegion Server 60010 60030
HQuorumPeer 2181
Sqoop Metastore 16000
If you have any concerns or need some changes in this doc, please contact us via below mail address.
ballicon@gmail.com
rbharath4u@gmail.com
saikiran.isk@gmail.com
email2anandraj@gmail.com
silambarasan909@gmail.com
End of Page 21/21

Hadoop Imp Commands

Uploaded by

Copyright:

Available Formats

Hadoop Imp Commands

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Hadoop Imp Commands

Uploaded by

Copyright:

Available Formats

Hadoop Hackers Groups hadoophackers@googlegroups.

If you are inside BIN folder of Hadoop Folder

To Start Hadoop (Present Working Directory should be Hadoop Folder)

To start HDFS Daemon process start-dfs.sh -> To Stop: stop-dfs.sh

Hadoop allows to start/stop each and every daemons by below commands.

Daemon Process Command

To Delete Name Node

To create Secure Key

End of Page 1/21

bin/hadoop dfs -copyFromLocal /home/administrator/R3.jar /ClientJavaF

From HDFS to Local

From HDFS to HDFS

From Local System to HDFS

From HDFS to HDFS

To list the files available

Note: between ls and / - space must be provided.

To Delete from HDFS

To Delete File inside Dir

To Rename the File name

To View Directory View

End of Page 2/21

To set Access permission

To run JAR file

To run Streaming-API or other program files apart from JAVA.

End of Page 3/21

[output file path]

hadoop jar /home/administrator/genericparserfiles.jar -files

Generic Parserer Used via HADOOP File Command

hadoop jar /home/administrator/genericparser.jar /hello.xt /GenericParsers/input2

Hadoop Safe Mode Command

End of Page 4/21

To load data in HIVE;

To write output in HDFS

insert overwrite local directory '/home/administrator/Desktop/Show' row format

End of Page 5/21

To Run User Defined Functions

To start HWI service

To Start Hive Server

To load data in Bucket:

End of Page 6/21

If you are inside BIN folder of HBase Folder

To Start Hbase (Present Working Directory should be Hbase Folder)

If you are inside BIN folder of Hbase Folder

To enter HBASE Prompt

To show tables use below comment

Note: Here QUERY

Below are some other methods to create table

End of Page 7/21

To get specific column

To get all columns available in table1

To get versions of tables

Description Command & Desc

End of Page 8/21

To Create & Load Data without Schema;

To Create & Load Data with Schema(via HDFS);

To Fetch Specific Column Data

End of Page 9/21

To take Procedural Plan Data

To Store Data in HDFS