cluster_start_stop_sequence.md

Start

1. ZooKeeper
   1. zkServer.sh start
   2. hdfs zkfc –formatZK
2. JournalNode (Hadoop)
   1. hadoop-daemon.sh start journalnode
3. NameNode (Hadoop)
   1. hdfs namenode -format
   2. adoop-daemon.sh start namenode
   3. hdfs namenode -bootstrapStandby
   4. hadoop-daemon.sh start namenode
4. ZKFC (Hadoop)
   1. hadoop-daemon.sh start zkfc
5. DataNode (Hadoop) 
   1. hadoop-daemons.sh start datanode
6. Main ResourceManager/NodeManager (Hadoop)
   1. start-yarn.sh
7. Backup ResourceManager (Hadoop)
   1. yarn-daemon.sh start resourcemanager
8. MapReduce JobHistory (Hadoop)
   1. mr-jobhistory-daemon.sh start historyserver
9. Master Hmaster/HRegionServer (HBase)
   1. start-hbase.sh
10. Backup Hmaster (HBase)
    1. hbase-daemon.sh start master

Stop

1. Close backup HMaster
   1. hbase-daemon.sh stop master
2. Shut down the main HMaster and HRegionServer
   1. stop-hbase.sh
3. Close the history log service
   1. mr-jobhistory-daemon.sh stop historyserver
4. Close ZKFC
   1. hadoop-daemon.sh stop zkfc
5. Close YARN separately at RM2
   1. yarn-daemon.sh stop resourcemanager
6. Close YARN at RM1
   1. stop-yarn.sh
7. Shut down all DataNodes in the cluster
   1. hadoop-daemons.sh stop datanode
8. Close NameNode2
   1. hadoop-daemon.sh stop namenode
9. Close the remaining NameNode
   1. hadoop-daemon.sh stop namenode
10. Close the JournalNode cluster 1. hadoop-daemon.sh stop journalnode
11. Shut down the ZooKeeper cluster 1. zkServer.sh stop