前面两篇文章给大家介绍了zookeeper的集群搭建，本篇文章来介绍下zookeeper的客户端命令操作
Zookeeper01之介绍和安装环境准备
Zookeeper02之集群环境搭建
bin目录下常用的脚本解释
| 脚本 | 说明 |
|---|---|
| zkCleanup | 清理Zookeeper历史数据，包括食物日志文件和快照数据文件 |
| zkCli | Zookeeper的一个简易客户端 |
| zkEnv | 设置Zookeeper的环境变量 |
| zkServer | Zookeeper服务器的启动、停止、和重启脚本 |
客户端命令行操作
1.配置zookeeper的环境变量
为了执行命令方便而不用每次都进入到zookeeper的安装目录去启动命令，我们可以将zookeeper配置到系统环境变量中。方便每次的操作
vim /etc/profile
# zookeeper 配置
export ZOOKEEPER_HOME=/root/app/zookeeper-3.4.10
export PATH=$PATH:$ZOOKEEPER_HOME/bin
重新加载配置文件
source /etc/profile
2.连接zookeeper
连接当前节点上的zookeeper服务器
zkCli.sh
想要连接其他节点上的服务器
zkCli.sh -timeout 5000 -server hadoop-node03[:2181]
| 参数 | 说明 |
|---|---|
| -timeout | 连接超时时长 |
| -server | 需要连接的服务器地址默认端口是2181，可以省略 |
3.数据操作
3.1zookeeper的数据存储结构
- 层次化的目录结构，命名符合常规文件系统规范
- 每个节点在zookeeper中叫做znode，并且有一个唯一的路径标识
- 节点znode可以包含数据和子节点（但是EPHEMERAL类型的节点不能有子节点）
- 客户端应用可以在节点上设置监听器
3.2节点的类型
1).znode有两种类型：
短暂性（ephemeral）（断开连接自己删除）
持久性（persistent）（断开连接不删除）
2).znode有四种形式的目录节点（默认是persistent）如下
| 序号 | 节点类型 | 描述 |
|---|---|---|
| 1 | PERSISTENT | 持久节点 |
| 2 | PERSISTENT_SEQUENTIAL | 持久有序节点 |
| 3 | EPHEMERAL | 短暂节点 |
| 4 | EPHEMERAL_SEQUENTIAL | 短暂有序节点 |
创建znode时设置顺序标识，znode名称后会附加一个值，顺序号是一个单调递增的计数器，有父节点维护
在分布式系统中，顺序号可以被用于为所有的事件进行全局排序，这样客户端可以通过顺序号推断事件的顺序
3.3命令说明
zookeeper中的常用命令不是很多，我们可以通过help命令来查看
[zk: hadoop-node03(CONNECTED) 0] help
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
命令说明
| 序号 | 命令 | 说明 |
|---|---|---|
| 1 | connect | 在连接服务器的情况连接到其他节点上去 |
| 2 | close | 关闭和服务器的连接 |
| 3 | ls | 用于获取某个节点下的节点信息 |
| 4 | ls2 | ls2命令是ls命令的增强版，比ls命令多输出本节点信息 |
| 5 | create | 用于创建节点，其中-s为顺序节点，-e临时节点 |
| 6 | delete | 只能删除没有子节点的节点 |
| 7 | rmr | 可以删除有子节点的节点 |
| 8 | get | 用户获取节点信息 |
| 9 | stat | 查看节点的状态 |
| 10 | set | 设置节点的数据 |
| 11 | listquota | listquota命令用于显示配额 |
| 12 | setquota | setquota命令用于设置节点个数以及数据长度的配额 |
| 13 | delquota | 用于删除配额，-n为子节点个数，-b为节点数据长度,如：delquota –n 2 |
| 14 | history | 用于列出最近的命令历史，可以和redo配合使用 |
| 15 | redo | 命令用于再次执行某个命令 |
| 16 | addauth | 用于节点认证，使用方式：如addauth digest username:password |
| 17 | setAcl | setAcl命令用于设置节点Acl Acl由三部分构成： 1为scheme，2为user，3为permission，一般情况下表示为scheme: id :permissions |
| 18 | getAcl | 获取节点的Acl，如getAcl /node1 |
| 19 | sync | 用于强制同步，由于请求在半数以上的zk server上生效就表示此请求生效，那么就会有一些zk server上的数据是旧的。sync命令就是强制同步所有的更新操作 |
| 20 | printwatchers | 用于设置和显示监视状态，值为on或则off |
| 21 | quit | 退出客户端 |
3.4常用操作
ls命令
ls命令是用来查看某个节点下的子节点信息，zookeeper的根节点是’/’,在初始的时候’/'下是没有任何子节点的
[zk: hadoop-node01(CONNECTED) 5] ls
[zk: hadoop-node01(CONNECTED) 6]
创建几个节点再查看:
[zk: hadoop-node01(CONNECTED) 8] create /node1 node1
Created /node1
[zk: hadoop-node01(CONNECTED) 9] create /node2 node2
Created /node2
[zk: hadoop-node01(CONNECTED) 10] ls /
[node2, zookeeper, node1]
ls2
ls2命令是ls命令的增强版，比ls命令多输出本节点信息
[zk: hadoop-node01(CONNECTED) 11] ls2 /
[node2, zookeeper, node1]
cZxid = 0x0
ctime = Thu Jan 01 08:00:00 CST 1970
mZxid = 0x0
mtime = Thu Jan 01 08:00:00 CST 1970
pZxid = 0x30000000c
cversion = 1
dataVersion = 0
aclVersion = 0
ephemeralOwner = 0x0
dataLength = 0
numChildren = 3
节点的信息描述在下面的get命令中描述
get命令
get命令查看节点的信息
[zk: hadoop-node01(CONNECTED) 12] get /
cZxid = 0x0
ctime = Thu Jan 01 08:00:00 CST 1970
mZxid = 0x0
mtime = Thu Jan 01 08:00:00 CST 1970
pZxid = 0x30000000c
cversion = 1
dataVersion = 0
aclVersion = 0
ephemeralOwner = 0x0
dataLength = 0
numChildren = 3
节点信息说明
| 序号 | 属性 | 说明 |
|---|---|---|
| 1 | cZxid | Znode创建的事务id |
| 2 | ctime | 节点创建时间 |
| 3 | mZxid | Znode被修改的事务id，即每次对znode的修改都会更新mZxid |
| 4 | mtime | 节点最近一次更新的时间 |
| 5 | pZxid | 是与该节点的子节点（或该节点）的最近一次 创建 / 删除 的时间戳对应 |
| 6 | cversion | 子节点数据更新次数 |
| 7 | dataVersion | 本节点数据更新次数 |
| 8 | aclVersion | 节点ACL(授权信息)的更新次数 |
| 9 | ephemeralOwner | 如果该节点为临时节点,ephemeralOwner值表示与该节点绑定的session id. 如果该节点不是临时节点,ephemeralOwner值为0 |
| 10 | dataLength | 节点数据长度 |
| 11 | numChildren | 子节点个数 |
zxid特别说明
对于zk来说，每次的变化都会产生一个唯一的事务id,zxid（ZooKeeper Transaction Id）。通过zxid，可以确定更新操作的先后顺序。例如，如果zxid1小于zxid2，说明zxid1操作先于zxid2发生。需要指出的是，zxid对于整个zk都是唯一的，即使操作的是不同的znode。
create命令
create命令用户创建ZNode节点。
create [-s] [-e] path data acl
| 参数 | 说明 |
|---|---|
| -s | 有序 |
| -e | 短暂 |
| path | 节点路径及名称 |
| data | 节点保存的数据 |
| acl | 节点的权限 |
创建永久节点
[zk: hadoop-node01(CONNECTED) 17] create /node1/node01 node01
Created /node1/node01
[zk: hadoop-node01(CONNECTED) 19] get /node1/node01
node01
cZxid = 0x30000000d
ctime = Tue Feb 26 06:00:43 CST 2019
mZxid = 0x30000000d
mtime = Tue Feb 26 06:00:43 CST 2019
pZxid = 0x30000000d
cversion = 0
dataVersion = 0
aclVersion = 0
ephemeralOwner = 0x0
dataLength = 6
numChildren = 0
[zk: hadoop-node01(CONNECTED) 20] ls /node1
[node01]
创建永久有序节点
[zk: hadoop-node01(CONNECTED) 21] create /node1/node02 node02
Created /node1/node02
[zk: hadoop-node01(CONNECTED) 22] delete /node1/node02
[zk: hadoop-node01(CONNECTED) 23] create -s /node1/node02 node02
Created /node1/node020000000002
[zk: hadoop-node01(CONNECTED) 24] get /node1/node020000000002
node02
cZxid = 0x300000010
ctime = Tue Feb 26 06:02:44 CST 2019
mZxid = 0x300000010
mtime = Tue Feb 26 06:02:44 CST 2019
pZxid = 0x300000010
cversion = 0
dataVersion = 0
aclVersion = 0
ephemeralOwner = 0x0
dataLength = 6
numChildren = 0
永久节点即使我们断开连接后再次连接后也不会消失
[zk: hadoop-node01(CONNECTED) 25] close
2019-02-25 23:07:21,968 [myid:] - INFO [main:ZooKeeper@684] - Session: 0x16920a6e1e90000 closed
[zk: hadoop-node01(CLOSED) 26] 2019-02-25 23:07:21,968 [myid:] - INFO [main-EventThread:ClientCnxn$EventThread@519] - EventThread shut down for session: 0x16920a6e1e90000
[zk: hadoop-node01(CLOSED) 26] connect
2019-02-25 23:07:33,387 [myid:] - INFO [main:ZooKeeper@438] - Initiating client connection, connectString=hadoop-node01 sessionTimeout=5000 watcher=org.apache.zookeeper.ZooKeeper
Main$MyWatcher@383534aa[zk: hadoop-node01(CONNECTING) 27] 2019-02-25 23:07:33,401 [myid:] - INFO [main-SendThread(hadoop-node01:2181):ClientCnxn$SendThread@1032] - Opening socket connection to server h
adoop-node01/192.168.88.121:2181. Will not attempt to authenticate using SASL (unknown error)2019-02-25 23:07:33,406 [myid:] - INFO [main-SendThread(hadoop-node01:2181):ClientCnxn$SendThread@876] - Socket connection established to hadoop-node01/192.168.88.121:2181, initi
ating session2019-02-25 23:07:33,423 [myid:] - INFO [main-SendThread(hadoop-node01:2181):ClientCnxn$SendThread@1299] - Session establishment complete on server hadoop-node01/192.168.88.121:21
81, sessionid = 0x16920a6e1e90001, negotiated timeout = 5000
WATCHER::
WatchedEvent state:SyncConnected type:None path:null
[zk: hadoop-node01(CONNECTED) 27] ls /node1
[node020000000002, node01]
创建短暂节点
[zk: hadoop-node01(CONNECTED) 28] create -e /node1/node03 node03
Created /node1/node03
[zk: hadoop-node01(CONNECTED) 29] ls /node1
[node020000000002, node01, node03]
创建短暂有序节点
[zk: hadoop-node01(CONNECTED) 30] create -s -e /node1/node03 node04
Created /node1/node030000000004
[zk: hadoop-node01(CONNECTED) 31] ls /node1
[node020000000002, node01, node03, node030000000004]
[zk: hadoop-node01(CONNECTED) 32]
注意
- 短暂节点不能创建子节点
- 创建短暂节点的客户端断开连接后数据会被自动删除
[zk: hadoop-node01(CONNECTED) 31] ls /node1
[node020000000002, node01, node03, node030000000004]
[zk: hadoop-node01(CONNECTED) 32] create -e /node1/node03/aaa aaa
Ephemerals cannot have children: /node1/node03/aaa
[zk: hadoop-node01(CONNECTED) 33] close
2019-02-25 23:10:35,224 [myid:] - INFO [main:ZooKeeper@684] - Session: 0x16920a6e1e90001 closed
[zk: hadoop-node01(CLOSED) 34] 2019-02-25 23:10:35,224 [myid:] - INFO [main-EventThread:ClientCnxn$EventThread@519] - EventThread shut down for session: 0x16920a6e1e90001
[zk: hadoop-node01(CLOSED) 34] connect
2019-02-25 23:10:44,441 [myid:] - INFO [main:ZooKeeper@438] - Initiating client connection, connectString=hadoop-node01 sessionTimeout=5000 watcher=org.apache.zookeeper.ZooKeeper
Main$MyWatcher@6bc168e5[zk: hadoop-node01(CONNECTING) 35] 2019-02-25 23:10:44,446 [myid:] - INFO [main-SendThread(hadoop-node01:2181):ClientCnxn$SendThread@1032] - Opening socket connection to server h
adoop-node01/192.168.88.121:2181. Will not attempt to authenticate using SASL (unknown error)2019-02-25 23:10:44,448 [myid:] - INFO [main-SendThread(hadoop-node01:2181):ClientCnxn$SendThread@876] - Socket connection established to hadoop-node01/192.168.88.121:2181, initi
ating session2019-02-25 23:10:44,464 [myid:] - INFO [main-SendThread(hadoop-node01:2181):ClientCnxn$SendThread@1299] - Session establishment complete on server hadoop-node01/192.168.88.121:21
81, sessionid = 0x16920a6e1e90002, negotiated timeout = 5000
WATCHER::
WatchedEvent state:SyncConnected type:None path:null
[zk: hadoop-node01(CONNECTED) 35] ls /node1
[node020000000002, node01]
set命令
set命令用来修改节点存储的信息
[zk: hadoop-node01(CONNECTED) 39] set /node1/node01 bbbb
cZxid = 0x30000000d
ctime = Tue Feb 26 06:00:43 CST 2019
mZxid = 0x300000018
mtime = Tue Feb 26 06:10:45 CST 2019
pZxid = 0x30000000d
cversion = 0
dataVersion = 1
aclVersion = 0
ephemeralOwner = 0x0
dataLength = 4
numChildren = 0
[zk: hadoop-node01(CONNECTED) 40] get /node1/node01
bbbb
cZxid = 0x30000000d
ctime = Tue Feb 26 06:00:43 CST 2019
mZxid = 0x300000018
mtime = Tue Feb 26 06:10:45 CST 2019
pZxid = 0x30000000d
cversion = 0
dataVersion = 1
aclVersion = 0
ephemeralOwner = 0x0
dataLength = 4
numChildren = 0
delete命令
delete命令用来删除节点，但是不能删除非空的节点
[zk: hadoop-node01(CONNECTED) 41] delete /node1
Node not empty: /node1
[zk: hadoop-node01(CONNECTED) 42] create /node3 node3
Created /node3
[zk: hadoop-node01(CONNECTED) 43] delete /node3
[zk: hadoop-node01(CONNECTED) 44] ls /
[node2, zookeeper, node1]
rmr命令
rmr命令也是用来删除节点的，而且rmr命令可以删除非空的节点。
[zk: hadoop-node01(CONNECTED) 45] ls /node1
[node020000000002, node01]
[zk: hadoop-node01(CONNECTED) 46] rmr /node1
[zk: hadoop-node01(CONNECTED) 47] ls /
[node2, zookeeper]
stat命令
stat命令查看节点的状态信息
[zk: hadoop-node01(CONNECTED) 49] stat /node2
cZxid = 0x30000000c
ctime = Tue Feb 26 05:38:33 CST 2019
mZxid = 0x30000000c
mtime = Tue Feb 26 05:38:33 CST 2019
pZxid = 0x30000000c
cversion = 0
dataVersion = 0
aclVersion = 0
ephemeralOwner = 0x0
dataLength = 5
numChildren = 0
close命令
关闭和服务器的连接
[zk: hadoop-node01(CONNECTED) 50] close
2019-02-25 23:24:54,024 [myid:] - INFO [main:ZooKeeper@684] - Session: 0x16920a6e1e90002 closed
[zk: hadoop-node01(CLOSED) 51] 2019-02-25 23:24:54,025 [myid:] - INFO [main-EventThread:ClientCnxn$EventThread@519] - EventThread shut down for session: 0x16920a6e1e90002
[zk: hadoop-node01(CLOSED) 51]
connect命令
再次连接服务器
[zk: hadoop-node03(CONNECTED) 0] close
2019-02-25 23:25:29,318 [myid:] - INFO [main:ZooKeeper@684] - Session: 0x3692632cde90001 closed
[zk: hadoop-node03(CLOSED) 1] 2019-02-25 23:25:29,321 [myid:] - INFO [main-EventThread:ClientCnxn$EventThread@519] - EventThread shut down for session: 0x3692632cde90001
[zk: hadoop-node03(CLOSED) 1] connect hadoop-node02:2181
2019-02-25 23:25:42,251 [myid:] - INFO [main:ZooKeeper@438] - Initiating client connection, connectString=hadoop-node02:2181 sessionTimeout=5000 watcher=org.apache.zookeeper.ZooK
eeperMain$MyWatcher@5d22bbb7[zk: hadoop-node02:2181(CONNECTING) 2] 2019-02-25 23:25:42,263 [myid:] - INFO [main-SendThread(hadoop-node02:2181):ClientCnxn$SendThread@1032] - Opening socket connection to serv
er hadoop-node02/192.168.88.122:2181. Will not attempt to authenticate using SASL (unknown error)2019-02-25 23:25:42,275 [myid:] - INFO [main-SendThread(hadoop-node02:2181):ClientCnxn$SendThread@876] - Socket connection established to hadoop-node02/192.168.88.122:2181, initi
ating session2019-02-25 23:25:42,286 [myid:] - INFO [main-SendThread(hadoop-node02:2181):ClientCnxn$SendThread@1299] - Session establishment complete on server hadoop-node02/192.168.88.122:21
81, sessionid = 0x26924b577110003, negotiated timeout = 5000
WATCHER::
WatchedEvent state:SyncConnected type:None path:null
[zk: hadoop-node02:2181(CONNECTED) 2]
quit命令
退出客户端连接
[zk: hadoop-node03(CONNECTED) 0] quit
Quitting...
2019-02-25 23:26:25,437 [myid:] - INFO [main-EventThread:ClientCnxn$EventThread@519] - EventThread shut down for session: 0x3692632cde90002
2019-02-25 23:26:25,438 [myid:] - INFO [main:ZooKeeper@684] - Session: 0x3692632cde90002 closed
[root@hadoop-node02 ~]#
zookeeper中的事件监听
在刚开始的时候我们介绍过zookeeper的两个核心功能是存储和监听，通过上面的命令我们已经可以对数据做存储操作了，接下来看下监听操作
1.数据改变的监听
监听的是节点中的内容改变了就会触发，针对的是具体的某个节点的信息，如下:
设置对某个节点内容改变的监听
[zk: localhost:2181(CONNECTED) 0] get /node2 watch
node2
cZxid = 0x30000000c
ctime = Tue Feb 26 05:38:33 CST 2019
mZxid = 0x30000000c
mtime = Tue Feb 26 05:38:33 CST 2019
pZxid = 0x30000000c
cversion = 0
dataVersion = 0
aclVersion = 0
ephemeralOwner = 0x0
dataLength = 5
numChildren = 0
在其他节点上修改node2的内容观察效果
但是要注意监听效果只会触发一次！！！
2.子节点改变的监听
给node2节点绑定子节点改变的监听器
[zk: localhost:2181(CONNECTED) 2] ls /node2 watch
[]
然后在node2下创建新的子节点观察效果
同样此监听器也只会触发一次!!!
好了常用的命令就介绍到此，下篇文章介绍节点的监听