CODEWITHSUNDEEP
hadoophbase
Raj
Raj

Reputation: 1

Hbase installation in three node hadoop cluster

I have installed my hadoop three node cluster(master,slave1 and slave2). I would like to install Hbase fully distrubuted mode. I am think to install HBase Master and Zookeepr in my hadoop cluster MASTER machine(i.e Namenode), And Region Servers in SLAVE1 and SLAVE2(i.e Datanodes) machines. Is this correct approach ?

Sorry, This may be simple question but I am new to NoSQL systems and want to do this installations.

I really appreciate If someone able to share any reference document for ths installation.

Thanks in advance.

Upvotes: 0

Views: 1313

Answers (1)

Rajesh N
Rajesh N

Reputation: 2574

In order to configure hbase and zookeeper on three nodes, i.e., 1 master and 2 slave nodes, you will need to edit hbase-site.xml, regionservers, hbase-env.sh (found in $HBASE_HOME/conf) and zoo.cfg (found in $ZOOKEEPER_HOME/conf).

Let us name your master node as master and slave nodes as slave1 and slave2. Let us consider your hadoop, hbase and zoopeeper folders are in /usr/local/cluster/ folder. Change the following files:

1. hbase-site.xml:

<configuration>
  <property>
    <name>hbase.master</name>
    <value>master:60000</value>
  </property>
  <property>
    <name>hbase.rootdir</name>
    <value>hdfs://master:8020/hbase</value>
  </property>
  <property>
  <property>
    <name>hbase.cluster.distributed</name>
    <value>true</value>
  </property>
    <name>hbase.zookeeper.property.clientPort</name>
    <value>2181</value>
  </property>
  <property>
    <name>hbase.zookeeper.quorum</name>
    <value>slave1,slave2</value>
  </property>
  <property>
    <name>hbase.tmp.dir</name>
    <value>/usr/local/cluster/zk-tmp</value>
  </property>
</configuration>

2. hbase-env.sh:

--add these lines--

export JAVA_HOME=/usr/lib/jvm/default-java
export HBASE_HOME=/usr/local/cluster/hbase
export HADOOP_HOME=/usr/local/cluster/hadoop

--modify these lines--

export HBASE_PID_DIR=/usr/local/cluster/zk-tmp
export HBASE_MANAGES_ZK=false

3. regionservers:

(delete the localhost and add these lines if you just want your regionservers in slave1 and slave2 only)

slave1
slave2

4. zoo.cfg:

--modify these lines--

dataDir=/usr/local/cluster/zk-tmp

--add these lines(since you start zookeeper server on master node)--

server.0=master:2888:3888

5. etc/hosts:

Edit the /etc/hosts file and comment the line with 127.0.1.1 (to avoid loopback address problems)

--add these lines--

your-master-node-ip master
your-slave1-node-ip slave1
your-slave2-node-ip slave2

Note: Do steps 1 to 5 in master, slave1 and slave2 nodes.

6. Start zookeeper server in master node:

$ZOOKEEPER_HOME/bin/zkServer.sh start

7. Start hbase processes in master node:

$HBASE_HOME/bin/start-hbase.sh

8. Check your hbase and zookeeper processes: Results for jps command in each node should contain-

--master--

QuorumPeerMain
HMaster
HRegionServer

--slave1--

HRegionServer

--slave2--

HRegionServer

9. Stopping zookeeeper and hbase:

$ZOOKEEPER_HOME/bin/zkServer.sh start
$HBASE_HOME/bin/stop-hbase.sh

Upvotes: 3

Related Questions