Zookeeper Cluster construction under CentOS

Stand-alone mode:

1) First download zookeeper compression package, here to use zookeeper3.4.8 ....

wget http://mirror.bit.edu.cn/apache/zookeeper/zookeeper-3.4.8/zookeeper-3.4.8.tar.gz

　　

2) unzip

First create the folder, put it in the/user/zookeeper/folder

mkdir Zookeeper1

　　

mkdir Zookeeper2

　　

mkdir Zookeeper3

Create three folders to hold three instances, unzip to the target folder, locate the compressed package storage path, execute the following command

　　

TAR-ZXVF zookeeper-3.4.8.tar.gz-c/usr/zookeeper/zookeeper1

　　

TAR-ZXVF zookeeper-3.4.8.tar.gz-c/usr/zookeeper/zookeeper2

　　

TAR-ZXVF zookeeper-3.4.8.tar.gz-c/usr/zookeeper/zookeeper3

3) operating unit price environment

First to complete the zookeeper1 of this example, first

　　

Use the LL command to display all file details,

Drwxr-xr-x. 2 4096 Mar 17:27 bin-rw-rw-r--. 1 83235 Feb 6 11:46 build.xml-rw-rw-r--. 1 88625 Feb 6 11:46 changes.txtdrwxr-xr-x. 2 4096 Mar 17:21 confdrwxr-xr-x. 4096 6 11:46 contribdrwxr-xr-x. 2 4096 Feb 6 11:50 dist-mavendrwxr-xr-x. 6 4096 Feb 6 11:49 docs-rw-rw-r--. 1 1953 Feb 6 11:46 ivysettings.xml-rw-rw-r--. 1 3498 Feb 6 11:46 ivy.xmldrwxr-xr-x. 4 4096 Feb 6 11:49 lib-rw-rw-r--. 1 11938 Feb 6 11:46 license.txt-rw-rw-r--. 1 171 Feb 6 11:46 notice.txt-rw-rw-r--. 1 1770 Feb 6 11:46 readme_packaging.txt-rw-rw-r--. 1 1585 Feb 6 11:46 readme.txtdrwxr-xr-x. 5 4096 Feb 6 11:46 recipesdrwxr-xr-x. 8 4096 Feb 6 11:49 src-rw-rw-r--. 1 1360961 Feb 6 11:46 zookeeper-3.4.8.jar-rw-rw-r--. 1 819 Feb 6 11:50 zookeeper-3.4.8.jar.asc-rw-rw-r--. 1 6 11:46 Zookeeper-3.4.8.jar.md5-rw-rw-r--. 1 1000 6 11:46 ZOOKEEPER-3.4.8.JAR.SHA1 

CP under Configuration file,

MV Conf/zoo_sample.cfg Conf/zoo.cfg

Start ZK,

./bin/zdserver.sh Start

4) Using a Java Client connection zookeeper

./bin/zkcli.sh-server 127.0.0.1:2181

Then you can use a variety of commands, similar to the file manipulation commands, and the input help can see all the commands.

5) Close

./bin/zdserver.sh Stop

　　

Distributed model (Replicated mode)

Zookeeper clusters are generally referred to as zookeeper Ensemble, or quorum.

Because we only have a machine here, through different port number, to make a pseudo-cluster way, if you have more machines, you can install a ZK on each machine, configuration is similar.

The so-called " pseudo-distributed cluster " is the instance in which multiple zookeeper are launched in a single PC. " fully distributed cluster " is each PC that launches a zookeeper instance.

Zookeeper there is no obvious master/slave relationship, each node is a server, leader Hung, will immediately follower from the election of one out as leader.

Because there is no master-slave relationship, nor do you have to configure SSH login without password, each ZK server is self-initiated, with each other through the TCP port to Exchange data

The Zookeeper2,zookeeper3 configuration file is changed, because Zookeeper1 when we do a stand-alone, has been modified, under the Zookeeper2,zookeeper3 conf folder, execute

MV Zoo_sample.cfg Zoo.cfg

　　

Create an Environment Catalog

Switch to Var/zookeeper path, create a new three folder Zookeeper1,zookeeper2,zookeeper3 .... For storing PID

Mkdir/var/zookeeper/zookeeper1mkdir/var/zookeeper/zookeeper2mkdir/var/zookeeper/zookeeper3

Create three new files in Zookeeper1,zookeeper2,zookeeper3, respectively myID

~ echo "1" >/var/zookeeper/zookeeper1/myid~ echo "2" >/var/zookeeper/zookeeper2/myid~ echo "3" >/var/zookeeper /zookeeper3/myid

　Note that myID can only be used for numbers, because I experimented with the following, containing English words ZK started up, throwing exceptions

分别修改配置文件
修改：dataDir,clientPort
增加：集群的实例，server.X，”X”表示每个目录中的myid的值

Vi/usr/zookeeper/zookeeper1/conf/zoo.cfg

　　　　

 

# of milliseconds of each tickticktime= -# The number of ticks that the initial # synchronization phase can takeinitlimit=Ten# The number of ticks that can pass between # Sending a request and getting an acknowledgementsynclimit=5# The DirectorywhereThe snapshot isstored.# DoNot use/tmp forStorage,/tmp here isJust # example Sakes.datadir=/var/zookeeper/zookeeper1# the port at which the clients would connectclientport=2181server.1=127.0.0.1:2888:3888server.2=127.0.0.1:2889:3889server.3=127.0.0.1:2890:3890# The maximum number of client connections.# increase This ifYou need to handle more Clients#maxclientcnxns= -# # is sure to read the maintenance sections of the # Administrator guide before turning on autopurge.## http://zookeeper.apache.org/doc/current/zookeeperadmin.html#sc_maintenance# # of snapshots to retaininchDatadir#autopurge.snapretaincount=3# Purge Task intervalinchhours# Set to"0"To disable Auto purge feature#autopurge.purgeinterval=1

The following only the different parts, the other and zk1 the same

Vi/usr/zookeeper/zookeeper2/conf/zoo.cfg

　

Datadir=/var/zookeeper/zookeeper2

　

clientport=2182

Vi/usr/zookeeper/zookeeper3/conf/zoo.cfg

　　

Datadir=/var/zookeeper/zookeeper3

　

clientport=2182

Start each machine

/usr/zookeeper/zookeeper1/bin/zkserver.sh start/usr/zookeeper/zookeeper2/bin/zkserver.sh start/usr/zookeeper/ zookeeper3/bin/zkserver.sh start

　　

View startup status

CentOS Zookeeper Cluster setup