-
Notifications
You must be signed in to change notification settings - Fork 0
Setting up HBase to store views
Installing Apache HBase is pretty straightforward: you just have to download the latest version from the official website and untar it in a folder of your choice.
Just a few steps divide you from starting up HBase and running a few shell commands to test it.
-
Open
conf/hbase-site.xml
and insert
It's advisable to set PIG_HOME inside your bash and add PIG_HOME/bin to your path.<configuration> <property> <name>hbase.rootdir</name> <value>file:///DIRECTORY/hbase</value> </property> <property> <name>hbase.zookeeper.property.dataDir</name> <value>/DIRECTORY/zookeeper</value> </property> </configuration>
Replace DIRECTORY in the above with the path to the directory you would have HBase and ZooKeeper write their data.
-
Set HBASE_HOME variable inside your bash (I advice to set it into .bashrc file):
# Set HBASE_HOME export HBASE_HOME=/home/daniele/HBase/hbase-0.94.13 export PIG_CLASSPATH=$PIG_HOME/pig-0.12.0-withouthadoop.jar:$HBASE_HOME/hbase-0.94.13.jar:$HBASE_HOME/lib/*:$HADOOP_HOME/lib/*
You should adapt this code to your Pig and HBase versions. We're also assuming that you set HADOOP_HOME and JAVA_HOME properly.
Before actually starting up HBase you should start up Hadoop.
Then you just need to type ./bin/start-hbase.sh
inside your shell. The following will appear starting Master, logging to logs/hbase-user-master-example.org.out
and you can refer to the log to check for errors.
To start the interactive shell type ./bin/hbase shell
.
- Create a table with name 'test' and column family 'cf':
create 'test', 'cf'
- Insert a line with the command
put 'test', 'row1', 'cf:a', 'value1'
- Verify the inserted data by running a scan of the table as follows:
scan 'test'
- You can also retrieve a single row with the command
get 'test', 'row1'
- Disable and drop your table:
disable 'test'
drop 'test'
- Exit the shell by typing 'exit'.
To access the HBase status UI, type http://localhost:60010 in your browser. It just gives you an insight into the cluster's state and the tables it hosts.