Zookeeper技术介绍

 ZooKeeper是一个分布式的，开放源码的分布式应用程序协调服务，是Google的Chubby一个开源的实现，是Hadoop和Hbase的重要组件。它是一个为分布式应用提供一致性服务的软件，提供的功能包括：配置维护、域名服务、分布式同步、组服务等。

  ZooKeeper的目标就是封装好复杂易出错的关键服务，将简单易用的接口和性能高效、功能稳定的系统提供给用户。

  ZooKeeper包含一个简单的原语集提供Java和C的接口。

  ZooKeeper代码版本中，提供了分布式独享锁、选举、队列的接口，代码在zookeeper-3.4.3src ecipes。其中分布锁和队列有Java和C两个版本，选举只有Java版本。

主要作用是用来解决分布式应用中经常遇到的一些数据管理问题，如集群管理、统一命名管理、分布式配置管理、分布式消息队列、分布式锁、分布式通知协调

官方网站：https://zookeeper.apache.org/

zookeeper体系结构

  zk的集群中有一个Leader，有一个或多个follower，它们之间是相互通信的，然后还存在一个客户端用来访问zk服务器。

  服务器端具有fast fail特性，一旦主故障后，会选举从成为新的主，主从模式是目前最常见的模式。

zookeeper名字空间由节点NODE构成,类似于文件系统,其中各个节点相当于目录和文件,通过路径作为唯一标识,与文件系统不同的是,每个节点具有与之对应的数据内容,同时也可以具有子节点,它用来存储协调数据,如状态、配置、位置信息、每个节点存储的数据量很小，KB级别.

watches介绍

zk对节点的增、删、改、查都可以触发监听

watch事件是一次性触发器，当它监视的数据发生变化时，通知设置了该客户端

Zookeeper安装布署

四台服务器

zk-001  10.0.0.8

zk-002  10.0.0.9

zk-003  10.0.0.10

zk-client 10.0.0.100

 

下载安装JDK环境

tar zxf jdk-8u60-linux-x64.tar.gz

mv jdk1.8.0_60 /usr/local/jdk

export JAVA_HOME=/usr/local/jdk

export CLASSPATH=.:$JAVA_HOME/lib/dt.jar:$JAVA_HOME/lib/tools.jar

export PATH=$JAVA_HOME/bin:$PATH

source /etc/profile

java -version

[root@zk-001 ~]# java -version

java version "1.8.0_60"

Java(TM) SE Runtime Environment (build 1.8.0_60-b27)

Java HotSpot(TM) 64-Bit Server VM (build 25.60-b23, mixed mode)

下载安装zookeeper

wget http://www.apache.org/dist/zookeeper/zookeeper-3.4.6/zookeeper-3.4.6.tar.gz

tar zxf zookeeper-3.4.6.tar.gz

mv zookeeper-3.4.6 /usr/local/zookeeper

[root@zk-001 ~]# cd /usr/local/zookeeper/

配置环境变量

export ZOOKEEPER_HOME=/usr/local/zookeeper

export PATH=$ZOOKEEPER_HOME/bin:$PATH

source /etc/profile

Zookeeper的配置

三种配置模式

1、伪分布模式

2、完全分布式

3、独立模式

独立模式配置

[root@zk-001 zookeeper]# cd conf/

[root@zk-001 conf]# ll

total 12

-rw-rw-r-- 1 1000 1000  535 Feb 20  2014 configuration.xsl

-rw-rw-r-- 1 1000 1000 2161 Feb 20  2014 log4j.properties

-rw-rw-r-- 1 1000 1000  922 Feb 20  2014 zoo_sample.cfg

[root@zk-001 conf]# cp zoo_sample.cfg zoo.cfg

[root@zk-001 conf]#vim zoo.cfg

# The number of milliseconds of each tick

tickTime=2000   #心跳检查周期（单位毫秒）

# the directory where the snapshot is stored.

# do not use /tmp for storage, /tmp here is just

# example sakes.

dataDir=/tmp/zookeeper  #存放数据目录

# the port at which the clients will connect

clientPort=2181  #供客户端连接的端口

集群模式配置

tickTime=2000

initLimit=10

syncLimit=5

//server.n=host:port1:port2 

数字N必须是myid中的值，Myid文件位于dataDIR目录下，只能是一个n值

//port1:leader端口，作为leader时，供follower连接的端口

//port2:选举端口，选举leader时，供follower连接的端口

server.1=s1:2888:3888

server.2=s2:2888:3888

server.3=s3:2888:3888

zk处理集群故障的算法是2n+1,最好在奇数机器数布署，如果同台服务器配置多实例就要注意端口冲突问题，可以写成如下

server.1=s1:2888:3888

server.2=s2:2889:3889

server.3=s3:2887:3887

 

配置过程如下

配置hosts文件

vim /etc/hosts

10.0.0.8 zk-001

10.0.0.9 zk-002

10.0.0.10 zk-003               

配置各服务器配置文件

[root@zk-001 conf]# mkdir /tmp/zookeeper

[root@zk-001 conf]# cd /tmp/zookeeper/

[root@zk-001 zookeeper]# echo 8 >myid

[root@zk-001 conf]# egrep -v "^#|^$" zoo.cfg

tickTime=2000

initLimit=10

syncLimit=5

dataDir=/tmp/zookeeper

clientPort=2181

server.8=zk-001:2888:3888

server.9=zk-002:2888:3888

server.10=zk-003:2888:3888

[root@zk-002 ~]# mkdir /tmp/zookeeper

[root@zk-002 ~]# cd /tmp/zookeeper/

[root@zk-002 zookeeper]# echo 9 >myid

[root@zk-002 zookeeper]# ll

total 4

-rw-r--r-- 1 root root 2 May 13 15:51 myid

[root@zk-002 zookeeper]# cd /usr/local/zookeeper/conf/

[root@zk-002 conf]# cp zoo_sample.cfg zoo.cfg

[root@zk-002 conf]# vim zoo.cfg

tickTime=2000

initLimit=10

syncLimit=5

dataDir=/tmp/zookeeper

clientPort=2181

server.8=zk-001:2888:3888

server.9=zk-002:2888:3888

server.10=zk-003:2888:3888

[root@zk-003 ~]# mkdir /tmp/zookeeper

[root@zk-003 ~]# cd /tmp/zookeeper/

[root@zk-003 zookeeper]# echo 10 >myid

[root@zk-003 zookeeper]# cd /usr/local/zookeeper/conf/

[root@zk-003 conf]# cp zoo_sample.cfg zoo.cfg

[root@zk-003 conf]# vim zoo.cfg

tickTime=2000

initLimit=10

syncLimit=5

dataDir=/tmp/zookeeper

clientPort=2181

server.8=zk-001:2888:3888

server.9=zk-002:2888:3888

server.10=zk-003:2888:3888

启动服务

[root@zk-001 bin]# zkServer.sh start

JMX enabled by default

Using config: /usr/local/zookeeper/bin/../conf/zoo.cfg

Starting zookeeper ... STARTED

[root@zk-002 bin]# zkServer.sh start

JMX enabled by default

Using config: /usr/local/zookeeper/bin/../conf/zoo.cfg

Starting zookeeper ... STARTED

[root@zk-003 bin]# zkServer.sh start

JMX enabled by default

Using config: /usr/local/zookeeper/bin/../conf/zoo.cfg

Starting zookeeper ... STARTED

查看状态

[root@zk-001 conf]# zkServer.sh status

JMX enabled by default

Using config: /usr/local/zookeeper/bin/../conf/zoo.cfg

Mode: follower

[root@zk-002 conf]# zkServer.sh status

JMX enabled by default

Using config: /usr/local/zookeeper/bin/../conf/zoo.cfg

Mode: follower

[root@zk-003 conf]# zkServer.sh status

JMX enabled by default

Using config: /usr/local/zookeeper/bin/../conf/zoo.cfg

Mode: leader

可以看出已经选举leader出来了，表明配置是正确的

测试leader故障后切换情况

[root@zk-003 conf]# jps

2694 Jps

2430 QuorumPeerMain

[root@zk-003 conf]# kill 2430

[root@zk-003 conf]# jps     

2724 Jps

[root@zk-002 conf]# zkServer.sh status

JMX enabled by default

Using config: /usr/local/zookeeper/bin/../conf/zoo.cfg

Mode: follower

[root@zk-001 conf]# zkServer.sh status

JMX enabled by default

Using config: /usr/local/zookeeper/bin/../conf/zoo.cfg

Mode: leader

可以看出zk-001这台服务器被选举成新的leader了

Zookeeper客户端四字符指令

conf 配置信息

cons 连接信息

dump 未处理会话节点

envi 环境信息

reqs 未处理请求   

stat 统计信息

wchs 服务器watch的详细信息

wchp 列出指定路径下服务器信息

[root@zk-003 ~]# echo conf | nc 10.0.0.8 2181

clientPort=2181

dataDir=/tmp/zookeeper/version-2

dataLogDir=/tmp/zookeeper/version-2

tickTime=2000

maxClientCnxns=60

minSessionTimeout=4000

maxSessionTimeout=40000

serverId=8

initLimit=10

syncLimit=5

electionAlg=3

electionPort=3888

quorumPort=2888

peerType=0

[root@zk-003 ~]# echo envi | nc 10.0.0.8 2181   

Environment:

zookeeper.version=3.4.6-1569965, built on 02/20/2014 09:09 GMT

host.name=zk-001

java.version=1.8.0_121

java.vendor=Oracle Corporation

java.home=/usr/lib/jvm/java-1.8.0-openjdk-1.8.0.121-0.b13.el6_8.x86_64/jre

java.class.path=/usr/local/zookeeper/bin/../build/classes:/usr/local/zookeeper/bin/../build/lib/*.jar:/usr/local/zookeeper/bin/../lib/slf4j-log4j12-1.6.1.jar:/usr/local/zookeeper/bin/../lib/slf4j-api-1.6.1.jar:/usr/local/zookeeper/bin/../lib/netty-3.7.0.Final.jar:/usr/local/zookeeper/bin/../lib/log4j-1.2.16.jar:/usr/local/zookeeper/bin/../lib/jline-0.9.94.jar:/usr/local/zookeeper/bin/../zookeeper-3.4.6.jar:/usr/local/zookeeper/bin/../src/java/lib/*.jar:/usr/local/zookeeper/bin/../conf:

java.library.path=/usr/java/packages/lib/amd64:/usr/lib64:/lib64:/lib:/usr/lib

java.io.tmpdir=/tmp

java.compiler=<NA>

os.name=Linux

os.arch=amd64

os.version=2.6.32-504.el6.x86_64

user.name=root

user.home=/root

user.dir=/usr/local/zookeeper/conf

[root@zk-003 ~]# echo stat | nc 10.0.0.8 2181       

Zookeeper version: 3.4.6-1569965, built on 02/20/2014 09:09 GMT

Clients:

 /10.0.0.10:41192[0](queued=0,recved=1,sent=0)

Latency min/avg/max: 0/0/0

Received: 4

Sent: 3

Connections: 1

Outstanding: 0

Zxid: 0x200000000

Mode: leader

Node count: 4

Zookeeper客户端命令

通过zkCli.sh 连接服务器

zkCli.sh -server zk-001:2181

         -server 10.0.0.9:2181

[root@zk-001 ~]# zkCli.sh -server zk-002:2181             

Connecting to zk-002:2181

WATCHER::

WatchedEvent state:SyncConnected type:None path:null

[zk: zk-002:2181(CONNECTED) 0] h ##帮助命令

ZooKeeper -server host:port cmd args

        stat path [watch]

        set path data [version]

        ls path [watch]

        delquota [-n|-b] path

        ls2 path [watch]

        setAcl path acl

        setquota -n|-b val path

        history

        redo cmdno

        printwatches on|off

        delete path [version]

        sync path

        listquota path

        rmr path

        get path [watch]

        create [-s] [-e] path data acl

        addauth scheme auth

        quit

        getAcl path

        close

        connect host:port

[zk: zk-002:2181(CONNECTED) 1] ls /             

#列出目录下内容

[zookeeper]

[zk: zk-002:2181(CONNECTED) 2] ls /zookeeper

[quota]

[zk: zk-002:2181(CONNECTED) 3] ls /zookeeper/quota

[]

[zk: zk-002:2181(CONNECTED) 4] create /root testfile  

#创建目录及数据

Created /root

[zk: zk-002:2181(CONNECTED)5]get /root #获取数据

testfile

cZxid = 0x200000002

ctime = Sat May 13 20:09:52 CST 2017

mZxid = 0x200000002

mtime = Sat May 13 20:09:52 CST 2017

pZxid = 0x200000002

cversion = 0

dataVersion = 0

aclVersion = 0

ephemeralOwner = 0x0

dataLength = 8

numChildren = 0

[zk: zk-002:2181(CONNECTED) 6] set /root testfile-001

cZxid = 0x200000002

ctime = Sat May 13 20:09:52 CST 2017

mZxid = 0x200000003

mtime = Sat May 13 21:04:17 CST 2017

pZxid = 0x200000002

cversion = 0

dataVersion = 1

aclVersion = 0

ephemeralOwner = 0x0

dataLength = 12

numChildren = 0

需要注意的是ZK是不能一次创建多级节点的

[zk: zk-002:2181(CONNECTED) 7] create /root/s1/s1-1

[zk: zk-002:2181(CONNECTED) 8] ls /root #提示是空

[]

[zk: zk-002:2181(CONNECTED) 9] create /root/s1 s1-data

Created /root/s1

[zk: zk-002:2181(CONNECTED) 10] create /root/s2 s2-data

Created /root/s2

[zk: zk-002:2181(CONNECTED) 11] create /root/s3 s3-data

Created /root/s3

[zk: zk-002:2181(CONNECTED) 13] ls /root

[s3, s1, s2]

查看状态

[zk: zk-002:2181(CONNECTED) 14] stat /root

cZxid = 0x200000002

ctime = Sat May 13 20:09:52 CST 2017

mZxid = 0x200000003

mtime = Sat May 13 21:04:17 CST 2017

pZxid = 0x200000006

cversion = 3

dataVersion = 1

aclVersion = 0

ephemeralOwner = 0x0

dataLength = 12

numChildren = 3

删除目录

[zk: zk-002:2181(CONNECTED) 15] delete /root/s3

[zk: zk-002:2181(CONNECTED) 16] ls /root

[s1, s2]

断开连接

[zk: zk-002:2181(CONNECTED) 17] close

2017-05-13 21:13:00,603 [myid:] - INFO[main:ZooKeeper@684] - Session: 0x95c0122efa50000 closed

2017-05-13 21:13:00,604 [myid:] - INFO[main-EventThread:ClientCnxn$EventThread@512] - EventThread shut down

重新连接

由于ZK服务器间的数据是一致的，因此这次我连接其它服务器

[zk: zk-002:2181(CLOSED) 18] connect zk-001:2181

[zk: zk-001:2181(CONNECTED) 19] get /root

testfile-001

cZxid = 0x200000002

ctime = Sat May 13 20:09:52 CST 2017

mZxid = 0x200000003

mtime = Sat May 13 21:04:17 CST 2017

pZxid = 0x200000007

cversion = 4

dataVersion = 1

aclVersion = 0

ephemeralOwner = 0x0

dataLength = 12

numChildren = 2

[zk: zk-001:2181(CONNECTED) 20] get /root/s1

s1-data

cZxid = 0x200000004

ctime = Sat May 13 21:07:42 CST 2017

mZxid = 0x200000004

mtime = Sat May 13 21:07:42 CST 2017

pZxid = 0x200000004

cversion = 0

dataVersion = 0

aclVersion = 0

ephemeralOwner = 0x0

dataLength = 7

numChildren = 0