虚拟机重启后hadoop走动步骤

/usr/local/zk/bin/zkServer.sh start




zkCli.sh

格式化ZooKeeper集群，目的是在ZooKeeper集群上建立HA的相应节点。

在hadoop1上执行命令：/usr/local/hadoop/bin/hdfs zkfc –formatZK【格式化操作的目的是在ZK集群中建立一个节点，用于保存集群c1中NameNode的状态数据】
在hadoop3上执行命令：/usr/local/hadoop/bin/hdfs zkfc –formatZK
【集群c2也格式化，产生一个新的ZK节点cluster2】

启动JournalNode集群

在hadoop1、hadoop2、hadoop3上分别执行命令/usr/local/hadoop/sbin/hadoop-daemon.sh start journalnode
【启动JournalNode后，会在本地磁盘产生一个目录，用户保存NameNode的edits文件的数据】

格式化集群c1的一个NameNode

从hadoop1和hadoop2中任选一个即可，这里选择的是hadoop1在hadoop1执行以下命令：/usr/local/hadoop/bin/hdfs namenode -format -clusterId c1【格式化NameNode会在磁盘产生一个目录，用于保存NameNode的fsimage、edits等文件】

启动c1中刚才格式化的NameNode

在hadoop1上执行命令：/usr/local/hadoop/sbin/hadoop-daemon.sh start namenode【启动后，产生一个新的java进程NameNode】

把NameNode的数据从hadoop1同步到hadoop2中

在hadoop2上执行命令：/usr/local/hadoop/bin/hdfs namenode –bootstrapStandby

启动c1中另一个Namenode

在hadoop2上执行命令：/usr/local/hadoop/sbin/hadoop-daemon.sh start namenode【产生java进程NameNode】

格式化集群c2的一个NameNode

从hadoop3和hadoop4中任选一个即可，这里选择的是hadoop3在hadoop3执行以下命令：/usr/local/hadoop/bin/hdfs namenode -format -clusterId c2

启动c2中刚才格式化的NameNode

在hadoop3上执行命令：/usr/local/hadoop/sbin/hadoop-daemon.sh start namenode

把NameNode的数据从hadoop3同步到hadoop4中

在hadoop4上执行命令：/usr/local/hadoop/bin/hdfs namenode –bootstrapStandby

.启动c2中另一个Namenode

在hadoop4上执行命令：/usr/local/hadoop/sbin/hadoop-daemon.sh start namenode

启动所有的DataNode

在hadoop1上执行命令：/usr/local/hadoop/sbin/hadoop-daemons.sh start datanode命令输出：[root@hadoop1 hadoop]# /usr/local/hadoop/sbin/hadoop-daemons.sh start datanode
hadoop1: starting datanode, logging to /usr/local/hadoop/logs/hadoop-root-datanode-hadoop1.out
hadoop3: starting datanode, logging to /usr/local/hadoop/logs/hadoop-root-datanode-hadoop3.out
hadoop2: starting datanode, logging to /usr/local/hadoop/logs/hadoop-root-datanode-hadoop2.out
hadoop4: starting datanode, logging to /usr/local/hadoop/logs/hadoop-root-datanode-hadoop4.out
[root@hadoop1 hadoop]#【上述命令会在四个节点分别启动DataNode进程】验证（以hadoop1为例）：[root@hadoop1 hadoop]# jps
23396 JournalNode
24302 Jps
24232 DataNode
23558 NameNode
22491 QuorumPeerMain
[root@hadoop1 hadoop]#

启动Yarn

在hadoop1上执行命令：/usr/local/hadoop/sbin/start-yarn.sh命令输出：[root@hadoop1 hadoop]# /usr/local/hadoop/sbin/start-yarn.sh
starting yarn daemons
starting resourcemanager, logging to /usr/local/hadoop/logs/yarn-root-resourcemanager-hadoop1.out
hadoop4: starting nodemanager, logging to /usr/local/hadoop/logs/yarn-root-nodemanager-hadoop4.out
hadoop3: starting nodemanager, logging to /usr/local/hadoop/logs/yarn-root-nodemanager-hadoop3.out
hadoop2: starting nodemanager, logging to /usr/local/hadoop/logs/yarn-root-nodemanager-hadoop2.out
hadoop1: starting nodemanager, logging to /usr/local/hadoop/logs/yarn-root-nodemanager-hadoop1.out
[root@hadoop1 hadoop]#验证：[root@hadoop1 hadoop]# jps
23396 JournalNode
25154 ResourceManager
25247 NodeManager
24232 DataNode
23558 NameNode
22491 QuorumPeerMain
25281 Jps
[root@hadoop1 hadoop]#【产生java进程ResourceManager和NodeManager】也可以通过浏览器访问，如下图

启动ZooKeeperFailoverController

在hadoop1、hadoop2、hadoop3、hadoop4上分别执行命令：/usr/local/hadoop/sbin/hadoop-daemon.sh start zkfc命令输出（以hadoop1为例）：[root@hadoop1 hadoop]# /usr/local/hadoop/sbin/hadoop-daemon.sh start zkfc
starting zkfc, logging to /usr/local/hadoop/logs/hadoop-root-zkfc-hadoop101.out
[root@hadoop1 hadoop]#验证（以hadoop1为例）：[root@hadoop1 hadoop]# jps
24599 DFSZKFailoverController
23396 JournalNode
24232 DataNode
23558 NameNode
22491 QuorumPeerMain
24654 Jps
[root@hadoop1 hadoop]#【产生java进程DFSZKFailoverController】

hadoop1,hadoop2.hadoop3.hadoop4:/usr/local/hadoop/sbin/hadoop-daemon.sh stop zkfc
hadoop1:/usr/local/hadoop/sbin/stop-yarn.sh
hadoop1: /usr/local/hadoop/sbin/hadoop-daemons.sh sotp datanode
hadoop1,hadoop2.hadoop3.hadoop4: /usr/local/hadoop/sbin/hadoop-daemon.sh stop namenode
hadoop1,hadoop2.hadoop3:/usr/local/hadoop/sbin/hadoop-daemon.sh stop journalnode
hadoop1,hadoop2.hadoop3:/usr/local/zk/bin/zkServer.sh stop