Build (28) zookeeper cluster

3 to build a cluster of machines that zookeeper to be installed in three hosts, each host is to be installed jdk.

1. Upload the file and extract the zookeeper

2. modify environment variables

vi / etc / profile

Add Content:

export ZOOKEEPER_HOME=/home/hadoop/zookeeper-3.4.5

export PATH=$PATH:$ZOOKEEPER_HOME/bin

Recompile the file:

source /etc/profile

Note: 3 sets zookeeper will need to be modified

3. Modify the zookeeper's profile

 

cd zookeeper-3.4.5/conf

 

cp zoo_sample.cfg zoo.cfg # copy zoo_sample.cfg , named zoo.cfg

 

we zoo.cfg

Zoo.cfg file contents are as follows:

 

tickTime: heartbeat cycle initLimit: How many heartbeats time it takes to initialize   

syncLimit: a request to obtain the maximum time difference between the response, if the difference exceeds the request is not considered the other's response, the other party may hang up

dataDir: data directory, save the data clientPort: zookeeper client access port

 

 

 

 

 

Add Content:

 

dataDir=/home/hadoop/zookeeper-3.4.5/data

dataLogDir=/home/hadoop/zookeeper-3.4.5/log

server.1 = slave1: 2888: 3888 ( server. Server id = hostname / host ip address : heartbeat port: data port )

server.2=slave2:2888:3888

server.3=slave3:2888:3888

Port communication between the leader and follower: # 2888

# 3888: communication port when the vote for leader

 

4. Create a data directory and log directory

cd /home/hadoop/zookeeper-3.4.5/

mkdir -m 755 data

mkdir -m 755 log

5. The id data recorded to the directory

3 machines in the data directory under the new myid file, myid the machine as the present content file node id zookeeper

For the first:

 

cd data

 

vi myid

 

Add Content

 

1

 

Or directly echo 1> myid

The other two similar

6. Cluster issued to the other machine

scp -r /home/hadoop/zookeeper-3.4.5 hadoop@slave2:/home/hadoop/ 

# The /home/hadoop/zookeeper-3.4.5 directory hadoop nominal copied to the slave host / home / hadoop directory

scp -r /home/hadoop/zookeeper-3.4.5 hadoop@slave3:/home/hadoop/

7. Turn off the firewall

Clusters are generally not exposed to the outside world, so the three machines are closed all ports firewall

service iptables stop

8. Start on all machines zookeeper

Enter zookeeper-3.4.5 / bin

zkServer.sh start

9. Check the cluster status

(1) JPS (see the process)

(2) zkServer.sh Status (see cluster status information from the master , the first to enter the ZooKeeper-3.4.5 / bin , if the cluster is not functioning properly, there will be error in this order )

 

Guess you like

Origin www.cnblogs.com/paradis/p/11373908.html