…0

Hbase Lab Manual

1. General commands
In Hbase, general commands are categorized into following commands
1. Status hbase(main):001:0> status /version/WHOAMI/table_help
2. Version
3. table_help ( scan, drop, get, put, disable, etc.)
4. Whoami

2. Tables Managements commands

These commands will allow programmers to create tables and table schemas with rows and column
families.The following are Table Management commands
1. Create
2. List
3. Describe
4. Disable
5. Disable_all
6. Enable
7. Enable_all
8. Drop
9. Drop_all
10. Show_filters
11. Alter
12. Alter_status

1.Create- To Create a table under Hbase

Syntax: create <tablename>, <columnfamilyname>
Ex- hbase(main):007:0> create 'sunil','personaldetail','educationaldetails'
lIVE: create 'cdac1','course','type'
2.List - Will display all the tables that are present or created in HBase
Syntax: list
3.Describe will give more information about column families present in the mentioned table
Syntax: describe <table name>
hbase(main):010:0> describe 'cdac' / TABLE NAME IS "cdac"
4.disable This command will start disabling the named table
If table needs to be deleted or dropped, it has to disable first
Syntax: disable <tablename>
hbase(main):011:0> disable 'education'

5.disable_all This command will disable all the tables matching the given regex.
The implementation is same as delete command (Except adding regex for matching)
Once the table gets disable the user can able to delete the table from HBase
Before delete or dropping table, it should be disabled first
Syntax: disable_all<"matching regex"

6.Enable This command will start enabling the named table

Whichever table is disabled, to retrieve back to its previous state we use this command
Syntax: enable <tablename>

7.show_filtersThis command displays all the filters present in HBase like ColumnPrefix Filter,
TimestampsFilter, PageFilter, FamilyFilter, etc.
Syntax: show_filters
8.drop To delete/DROP the table present in HBase, first we have to disable it.
So either table to drop or delete first the table should be disable using disable command
Syntax: drop <table name>
9.drop_all This command will drop all the tables matching the given regex
Syntax: drop_all<"regex">
10.is_enabled This command will verify whether the named table is enabled or not.
Suppose a table is disabled, to use that table we have to enable it by using enable command
is_enabled command will check either the table is enabled or not
Syntax: is_enabled 'education'
11..alter Changing the Maximum Number of Cells of a Column Family
Syntax: alter <tablename>, NAME=><column familyname>, VERSIONS=>5
Altering single, multiple column family names
Syntax: alter ‘<table_name>’, ‘reference of column_family’ , { NAME => 'New _column_family
name', IN_MEMORY => true, VERSIONS => 5}
Ex- alter 'student','Personal_details', {NAME => 'Admission_details', IN_MEMORY => true,
VERSIONS => 5}
Before alter command table structure in Hbase
Personal_details Education_details
Name Age Address Course Year Grade
After alter command
Personal_details Admission_details Education_details
Name Age Address Course Year Grade
 Deleting column family names from table
alter 'sun', 'delete' =>'weather' // only column family
alter 'education', 'delete' =>’waether'
12. alter_status - can get the status of the alter command

Which indicates the number of regions of the table that have received the updated schema
pass table name
Syntax: alter_status 'education'(Table Name)

3.Data manipulation commands

These commands will work on the table related to data manipulations such as putting data into a table,
retrieving data from a table and deleting schema, etc.
The commands come under these are
1. Count
2. Put
3. Get
4. Delete
5. Delete all
6. Truncate
7. Scan
1.Count The command will retrieve the count of a number of rows in a table. The value returned by
this one is the number of rows.
Syntax: count <'tablename'>, CACHE =>1000

Ex- hbase> count 'guru99', CACHE=>1000

Current count is shown per every 1000 rows by default.

Count interval may be optionally specified.
Default cache size is 10 rows.

2. Put - It will put a cell ‘value’ at defined or specified table or row or column. It will optionally
coordinate time stamp
Syntax: put <'tablename'>,<'rowname'>,<'columnvalue'>,<'value'> OR

put '<HBase_table_name>', 'row_key', '<colfamily:colname>', '<value>'

eX- hbase(main):018:0> put 'cdac1',1,'course:Bridgecourse','type:paid'

hbase(main):018:0>put 'cdac1',1,'course:online','type:free'

hbase(main):020:0> scan 'cdac1'

Example: Here we are placing values into table “guru99” under row r1 and column c1
hbase> put 'guru99', 'r1', 'c1', 'value', 10 /where r1 is row and c1 is column
3.Get- By using this command, you will get a row or cell contents present in the table.
Syntax: get <'tablename'>, <'rowname'>, {< Additional parameters>}
 eX-hbase> get 'guru99', 'r1', {COLUMN => 'c1'}

4.Delete This command will delete cell value at defined table of row or column.

Syntax:delete <'tablename'>,<'row name'>,<'column name'>

Delete must and should match the deleted cells coordinates exactly.
When scanning, delete cell suppresses older versions of values.
eX-hbase(main):)020:0> delete 'guru99', 'r1', 'c1''.
The above execution will delete row r1 from column family c1 in table “guru99.”
5.deleteall This Command will delete all cells in a given row.
Syntax: deleteall <'tablename'>, <'rowname'>

eX- hbase>deleteall 'guru99', 'r1', 'c1'

6.Truncate
Syntax: truncate <tablename>
After truncate of an hbase table, the schema will present but not the records. This command
performs 3 functions; those are listed below
Disables table if it already presents
Drops table if it already presents
Recreates the mentioned table
7.Scan- Display the Content of HBase Table
Syntax: scan <'tablename'>, {Optional parameters}
Scanner specifications may include one or more of the following attributes.
These are TIMERANGE, FILTER, TIMESTAMP, LIMIT, MAXLENGTH, COLUMNS, CACHE, STARTROW and
STOPROW.
eX-scan 'guru99'
Create a table who has more than 01 column families-

1. There are two column families CF1 and CF2 in creating table Hbase
*you can only add one column and one column, not multiple columns at the same time.
This table has two column families, CF1 and CF2. Under CF1 and CF2, there are two columns, name
and gender, Chinese and Math
*. {NAME=>'cf1'} / name should be in capital i.e NAME

Example:
hbase(main):041:0> create 'hbase_1102', {NAME=>'cf1'}, {NAME=>'cf2'}
2. Add data to the table. When you want to add data to the table of HBase,
hbase(main):042:0> put'hbase_1102', '001','cf1:name','Sumit'
hbase(main):043:0> put'hbase_1102', '001','cf1:gender','male'
 hbase(main):044:0> put'hbase_1102', '001','cf2:chinese','90'
hbase(main):045:0> put'hbase_1102', '001','cf2:math','91'

To Create a table with 03 colun family in Hbase

1: create 'student', {NAME=>'Academic'}, {NAME=>'Personal'},{NAME=>'placement'}
A.insert data under Academic column family
a. hbase(main):4:0>put'student', '001','Academic:Grade','First'
b. hbase(main):5:0>put'student', '001','Academic:Attendance','Good'
c. hbase(main):6:0>put'student', '001','Academic:Eassywriting','Average'
d. hbase(main):7:0>put'student', '001','Academic:Computerproficency','Yes'
scan 'student' to check all the fields data inserted
B.insert data under Personal column family
a. hbase(main):010:0> put'student', '001','Personal:Name','Tarun'
b. hbase(main):011:0> put'student', '001','Personal:mobno','8589658965'
c. hbase(main):012:0> put'student', '001','Personal:height','5'

C.insert data under Placement column family

a. hbase(main):004:0> put 'student','001','placement:Company','HCL'
b. hbase(main):005:0> put 'student','001','placement:Year','2018'
c. hbase(main):006:0> put 'student','001','placement:Package','6 Lakh P.A'

output of the Table:

Student:
Academic Personal Placement

R Grade Attendance Essay Computer Name Mob Height Weight Company Year Package
o writing No.
proficency
w
1 First Good Average Yes Tarun 8589 5" 58 HCL 2018 6 LPA

hbase(main):007:0> scan 'student'

ROW COLUMN+CELL
001 column=Academic:Attendance, timestamp=1660106971926, value=Good
001 column=Academic:Computerproficency, timestamp=1660107016304, value=Yes
001 column=Academic:Eassywriting, timestamp=1660106995630, value=Average
001 column=Academic:Grade, timestamp=1660106942707, value=First
001 column=Personal:Name, timestamp=1660107080883, value=Tarun
001 column=Personal:height, timestamp=1660107113454, value=5
001 column=Personal:mobno, timestamp=1660107098959, value=8589658965
001 column=placement:Company, timestamp=1660107453461, value=HCL
001 column=placement:Package, timestamp=1660107499919, value=6 Lakh P.A
001 column=placement:Year, timestamp=1660107478876, value=2018

4. Cluster Replication Commands

These commands work on cluster set up mode of HBase.
For adding and removing peers to cluster and to start and stop replication these commands are used
in general.
Command Functionality
Add peers to cluster to replicate
add_peer
hbase> add_peer ‘3’, zk1,zk2,zk3:2182:/hbase-prod
Stops the defined replication stream.
Deletes all the metadata information about the
remove_peer
peer
hbase> remove_peer ‘1’
Restarts all the replication features
start_replication
hbase> start_replication
Stops all the replication features
stop_replication
hbase>stop_replication

HBase architecture always has “Single Point Of Failure” feature, and there is no exception handling
mechanism associated with it
Performance Bottlenecks in HBase

In any production environment, HBase is running with a cluster of more than 5000 nodes, only
Hmaster acts as the master to all the slaves Region servers. If Hmaster goes down, it can be only be
recovered after a long time. Even though the client is able to connect region server. Having another
master is possible but only one will be active. It will take a long time to activate the second Hmaster if
the main Hmaster goes down. So, Hmaster is a performance bottleneck.
In HBase, we cannot implement any cross data operations and joining operations, of course, we can
implement the joining operations using MapReduce, which would take a lot of time to designing and
development. Tables join operations are difficult to perform in HBase. In some use case, its impossible
to create join operations that related to tables that are present in HBase
HBase would require new design when we want to migrate data from RDBMS external sources to
 HBase servers. However, this process takes a lot of time.
HBase is really tough for querying. We may have to integrate HBase with someSQL layers
like Apache phoenix where we can write queries to trigger the data in the HBase. It’s really good to
have Apache Phoenix on top of HBase.
Another drawback with HBase is that, we cannot have more than one indexing in the
table, only row key column acts as a primary key. So, the performance would be slow
when we wanted to search on more than one field or other than Row key. This problem
we can overcome by writing MapReduce code, integrating with Apache SOLR and with
Apache Phoenix.
Slow improvements in the security for the different users to access the data from HBase.
HBase doesn’t support partial keys completely
HBase allows only one default sort per table
It’s very difficult to store large size of binary files in HBase
The storage of HBase will limit real-time queries and sorting
Key lookup and Range lookup in terms of searching table contents using key values, it will limit
queries that perform on real time
Default indexing is not present in HBase. Programmers have to define several lines of code or script
to perform indexing functionality in HBase
Expensive in terms of Hardware requirements and memory blocks allocations.
More servers should be installed for distributed cluster environments (like each server for
NameNode, DataNodes, ZooKeeper, and Region Servers)
Performance wise it require high memory machines
Costing and maintenance wise it is also higher

Advantages of HBase
Here, we will learn what are the pros/benefits of HBase:
Can store large data sets on top of HDFS file storage and will aggregate and analyze billions of rows
present in the HBase tables
In HBase, the database can be shared
Operations such as data reading and processing will take small amount of time as compared to
traditional relational models
Random read and write operations
For online analytical operations, HBase is used extensively.
For example: In banking applications such as real-time data updates in ATM machines, HBase can be
used.

Here are the important cons/limitations of HBase:

We cannot expect completely to use HBase as a replacement for traditional models. Some of the
traditional models features cannot support by HBase
HBase cannot perform functions like SQL. It doesn’t support SQL structure, so it does not contain any
query optimizer
HBase is CPU and Memory intensive with large sequential input or output access while as Map
Reduce jobs are primarily input or output bound with fixed memory. HBase integrated with Map-
reduce jobs will result in unpredictable latencies
HBase integrated with pig and Hive jobs results in some time memory issues on cluster
In a shared cluster environment, the set up requires fewer task slots per node to allocate for HBase
CPU