解决zookeeper集群重启 Error contacting service. It is probably not running 问题

一 问题出现情形

1 集群启动后异常关闭，重启出错

2 集群未正常启动

二 查看zookeeper.out分析原因

 cat zookeeper.out

  1 网络问题。排查防火墙

1. java.net.NoRouteToHostException: 没有到主机的路由 (Host unreachable)

2. at java.net.PlainSocketImpl.socketConnect(Native Method)

3. at java.net.AbstractPlainSocketImpl.doConnect(AbstractPlainSocketImpl.java:350)

4. at java.net.AbstractPlainSocketImpl.connectToAddress(AbstractPlainSocketImpl.java:206)

5. at java.net.AbstractPlainSocketImpl.connect(AbstractPlainSocketImpl.java:188)

6. at java.net.SocksSocketImpl.connect(SocksSocketImpl.java:392)

7. at java.net.Socket.connect(Socket.java:589)

8. at org.apache.zookeeper.server.quorum.QuorumCnxManager.connectOne(QuorumCnxManager.java:558)

9. at org.apache.zookeeper.server.quorum.QuorumCnxManager.connectAll(QuorumCnxManager.java:610)

10. at org.apache.zookeeper.server.quorum.FastLeaderElection.lookForLeader(FastLeaderElection.java:838)

11. at org.apache.zookeeper.server.quorum.QuorumPeer.run(QuorumPeer.java:957)

解决步骤：问题排查 7 ->[2、6] ->重启集群

 

2 初次启动。

解决步骤：问题排查1-->2-->3-->4-->5-->7-->重启集群

备注：初次启动 个个环节都有可能有问题 如果有报错信息可以针对性解决

 

3 重新启动。

解决步骤：暂停集群-->2-->6-->7-->重启集群

备注：重新启动主要是，节点配置信息问题以及防火墙问题。需要按顺序，暂停zookeeper，kill进程，删除节点信息，关闭防火墙，再次启动集群

三 问题排查

1 java环境

java -version

2 排查端口占用(我就这里的问题)

netstat -apn | grep 2181   #默认2181端口为服务端提供端口

2020-08-14 13:45:39,162 [myid:1] - INFO  [QuorumPeer[myid=1]/0:0:0:0:0:0:0:0:2181:FastLeaderElection@813] - New election. My id =  1, proposed zxid=0x100000000
2020-08-14 13:45:39,162 [myid:1] - INFO  [WorkerReceiver[myid=1]:FastLeaderElection@595] - Notification: 1 (message format version), 1 (n.leader), 0x100000000 (n.zxid), 0x3 (n.round), LOOKING (n.state), 1 (n.sid), 0x2 (n.peerEpoch) LOOKING (my state)
2020-08-14 13:45:39,162 [myid:1] - INFO  [WorkerReceiver[myid=1]:FastLeaderElection@595] - Notification: 1 (message format version), 2 (n.leader), 0x200000000 (n.zxid), 0x3 (n.round), LOOKING (n.state), 2 (n.sid), 0x2 (n.peerEpoch) LOOKING (my state)
2020-08-14 13:45:39,162 [myid:1] - WARN  [WorkerSender[myid=1]:QuorumCnxManager@584] - Cannot open channel to 3 at election address /10.3.3.138:3888
java.net.ConnectException: 拒绝连接 (Connection refused)
        at java.net.PlainSocketImpl.socketConnect(Native Method)
        at java.net.AbstractPlainSocketImpl.doConnect(AbstractPlainSocketImpl.java:350)
        at java.net.AbstractPlainSocketImpl.connectToAddress(AbstractPlainSocketImpl.java:206)
        at java.net.AbstractPlainSocketImpl.connect(AbstractPlainSocketImpl.java:188)
        at java.net.SocksSocketImpl.connect(SocksSocketImpl.java:392)
        at java.net.Socket.connect(Socket.java:589)
        at org.apache.zookeeper.server.quorum.QuorumCnxManager.connectOne(QuorumCnxManager.java:558)
        at org.apache.zookeeper.server.quorum.QuorumCnxManager.toSend(QuorumCnxManager.java:534)
        at org.apache.zookeeper.server.quorum.FastLeaderElection$Messenger$WorkerSender.process(FastLeaderElection.java:454)
        at org.apache.zookeeper.server.quorum.FastLeaderElection$Messenger$WorkerSender.run(FastLeaderElection.java:435)
        at java.lang.Thread.run(Thread.java:748)
2020-08-14 13:45:39,163 [myid:1] - INFO  [WorkerSender[myid=1]:QuorumPeer$QuorumServer@184] - Resolved hostname: 10.3.3.138 to address: /10.3.3.138
2020-08-14 13:45:39,163 [myid:1] - INFO  [WorkerReceiver[myid=1]:FastLeaderElection@595] - Notification: 1 (message format version), 2 (n.leader), 0x200000000 (n.zxid), 0x3 (n.round), LOOKING (n.state), 1 (n.sid), 0x2 (n.peerEpoch) LOOKING (my state)
2020-08-14 13:45:39,163 [myid:1] - WARN  [WorkerSender[myid=1]:QuorumCnxManager@584] - Cannot open channel to 3 at election address /10.3.3.138:3888
java.net.ConnectException: 拒绝连接 (Connection refused)

 

备注：若集群未启动 则不应该有端口占用

kill -9 pid                #pid为占用端口的进程id号

3 排除网卡问题

ip addr

备注：如果为物理地址则 重启网卡

service network restart

4 排除网络问题

ping ip               #ping其他节点主机 若zoo.cfg 使用域名则用域名

备注：无ping命令则安装  yum install iputils-ping 若有问题则 host映射问题

 

5 排查节点配置信息

dataDir 是否存在 myid文件内容与 service.x 中x对应

例如：zk01  其对应service.x  则该目录下myid内容为1

6 删除节点残留信息

rm -rf version-2/ zookeeper_server.pid

7 防火墙拦截端口

systemctl status firewalld.service

1. systemctl stop firewalld.service #关闭防火墙

2. systemctl disable firewalld.service #禁止启动防火墙

 

本来以为是三台主机没有配置好，重装了好几次，最后总结了4个问题。

1：如果用阿里云服务器，链接，服务器没有防火墙，只有安全组，需要自己开放端口。（这次悲剧这里）

2：在本地进行连接的话 127.0.0.1配置注销了（很多博主都写了，我也不知道为啥）

3：记着设置为静态路由，动态的，修改比较麻烦。

4：zookeeper的原子广播协议（选举算法），将集群分为leader（有且只有一个），和follower，这些是从众多结点中随机选出，需要每一个结点的zookeeper都要开启，不然也会造成链接失败。（注：有的时候选举的特别慢，要耐心等候

下面几点需要注意的：

1、确认在每个$zookeeper_home/data/myid中有对应数字

2、是否关闭防火墙：systemctl stop firewalld，systemctl disable firewalld

3、zoo.cfg中的server需要写成以下形式的：