[[email protected]_71_178_centos zookeeper]# pwd
/usr/zookeeper

[[email protected]_71_178_centos zookeeper]# ls
s1  s2  s3  zk  zookeeper-3.4.10.tar.gz

tar -cvzf zookeeper-3.4.9.tar.gz -C zk 

[[email protected]_71_178_centos s1]# ls
zkdata  zklog  zookeeper-3.4.10

# The number of milliseconds of each tick
tickTime=2000
# The number of ticks that the initial
# synchronization phase can take
initLimit=10
# The number of ticks that can pass between
# sending a request and getting an acknowledgement
syncLimit=5
# the directory where the snapshot is stored.
# do not use /tmp for storage, /tmp here is just
# example sakes.
dataDir=/usr/zookeeper/s1/zkdata
dataLogDir=/usr/zookeeper/s1/zklog
# the port at which the clients will connect
clientPort=2181
# the maximum number of client connections.
# increase this if you need to handle more clients
#maxClientCnxns=60
#
# Be sure to read the maintenance section of the
# administrator guide before turning on autopurge.
#
# http://zookeeper.apache.org/doc/current/zookeeperAdmin.html#sc_maintenance
#
# The number of snapshots to retain in dataDir
#autopurge.snapRetainCount=3
# Purge task interval in hours
# Set to "0" to disable auto purge feature
#autopurge.purgeInterval=1
server.1=localhost:2287:3387
server.2=localhost:2288:3388
server.3=localhost:2289:3389
 

# The number of milliseconds of each tick
tickTime=2000
# The number of ticks that the initial
# synchronization phase can take
initLimit=10
# The number of ticks that can pass between
# sending a request and getting an acknowledgement
syncLimit=5
# the directory where the snapshot is stored.
# do not use /tmp for storage, /tmp here is just
# example sakes.
dataDir=/usr/zookeeper/s2/zkdata
dataLogDir=/usr/zookeeper/s2/zklog
# the port at which the clients will connect
clientPort=2182
# the maximum number of client connections.
# increase this if you need to handle more clients
#maxClientCnxns=60
#
# Be sure to read the maintenance section of the
# administrator guide before turning on autopurge.
#
# http://zookeeper.apache.org/doc/current/zookeeperAdmin.html#sc_maintenance
#
# The number of snapshots to retain in dataDir
#autopurge.snapRetainCount=3
# Purge task interval in hours
# Set to "0" to disable auto purge feature
#autopurge.purgeInterval=1
server.1=localhost:2287:3387
server.2=localhost:2288:3388
server.3=localhost:2289:3389
 

s3/zookeeper-3.4.10/conf/zook.cfg

# The number of milliseconds of each tick
tickTime=2000
# The number of ticks that the initial
# synchronization phase can take
initLimit=10
# The number of ticks that can pass between
# sending a request and getting an acknowledgement
syncLimit=5
# the directory where the snapshot is stored.
# do not use /tmp for storage, /tmp here is just
# example sakes.
dataDir=/usr/zookeeper/s3/zkdata
dataLogDir=/usr/zookeeper/s3/zklog
# the port at which the clients will connect
clientPort=2183
# the maximum number of client connections.
# increase this if you need to handle more clients
#maxClientCnxns=60
#
# Be sure to read the maintenance section of the
# administrator guide before turning on autopurge.
#
# http://zookeeper.apache.org/doc/current/zookeeperAdmin.html#sc_maintenance
#
# The number of snapshots to retain in dataDir
#autopurge.snapRetainCount=3
# Purge task interval in hours
# Set to "0" to disable auto purge feature
#autopurge.purgeInterval=1
server.1=localhost:2287:3387
server.2=localhost:2288:3388
server.3=localhost:2289:3389

• tickTime：這個時間是作為 Zookeeper 伺服器之間或客戶端與伺服器之間維持心跳的時間間隔，也就是每個 tickTime 時間就會發送一個心跳。
• dataDir：顧名思義就是 Zookeeper 儲存資料的目錄，預設情況下，Zookeeper 將寫資料的日誌檔案也儲存在這個目錄裡。
• clientPort：這個埠就是客戶端連線 Zookeeper 伺服器的埠，Zookeeper 會監聽這個埠，接受客戶端的訪問請求。

• initLimit：這個配置項是用來配置 Zookeeper 接受客戶端（這裡所說的客戶端不是使用者連線 Zookeeper 伺服器的客戶端，而是 Zookeeper 伺服器叢集中連線到 Leader 的 Follower 伺服器）初始化連線時最長能忍受多少個心跳時間間隔數。當已經超過 10 個心跳的時間（也就是 tickTime）長度後 Zookeeper 伺服器還沒有收到客戶端的返回資訊，那麼表明這個客戶端連線失敗。總的時間長度就是 5*2000=10 秒
• syncLimit：這個配置項標識 Leader 與 Follower 之間傳送訊息，請求和應答時間長度，最長不能超過多少個 tickTime 的時間長度，總的時間長度就是 2*2000=4 秒
• server.A=B：C：D：其中 A 是一個數字，表示這個是第幾號伺服器；B 是這個伺服器的 ip 地址；C 表示的是這個伺服器與叢集中的 Leader 伺服器交換資訊的埠；D 表示的是萬一叢集中的 Leader 伺服器掛了，需要一個埠來重新進行選舉，選出一個新的 Leader，而這個埠就是用來執行選舉時伺服器相互通訊的埠。如果是偽叢集的配置方式，由於 B 都是一樣，所以不同的 Zookeeper 例項通訊埠號不能一樣，所以要給它們分配不同的埠號。

除了修改 zoo.cfg 配置檔案，叢集模式下還要配置一個檔案 myid，這個檔案在 dataDir 目錄下，這個檔案裡面就有一個數據就是 A 的值，Zookeeper 啟動時會讀取這個檔案，拿到裡面的資料與 zoo.cfg 裡面的配置資訊比較從而判斷到底是那個 server。

1. 分別在資料夾zkdata中建立檔案myid檔案內容為server.?=localhost:2287:3387中的?如下：s1資料夾下vi /usr/zookeeper/s1/zkdata/myid

2. 1
   

3. 分別進入相應zookeeper中bin目錄分別啟動

4. [[email protected]_71_178_centos bin]# ./zkServer.sh  start

5. 查詢狀態是否成功

6. [[email protected]_71_178_centos bin]# ./zkServer.sh status
   ZooKeeper JMX enabled by default
   Using config: /usr/zookeeper/s1/zookeeper-3.4.10/bin/../conf/zoo.cfg
   Mode: follower
   

   幾點問題：1。zook.cfg中dataDir路徑一定要正確否則啟動不成功，2.myid中一定要跟server.幾對應否則也不成功
   為啥需要奇數機器：

   官網強烈建議部署zookeeper機器的數量是奇數的,原因在於ZooKeeper的選舉機制,選票超過半數的機器才有可能成為Leader,因此當整個叢集中只有兩臺伺服器或者整個叢集超過半數的機器都掛掉時,是無法選舉出Leader的.其實這是一個叢集資源利用最大化的問題,例如叢集中有三個節點,允許掛掉一個節點,這時還剩餘兩個節點,超過半數;而叢集中有四個節點時,最多也只能掛掉一個節點,因為如果再掛掉一個,整個叢集就剩下兩個節點,無法滿足超過叢集數的半數這個條件.綜合來看兩者的容災能力是一樣的