2, zookeeper: zoo. CFG, data folder, modify myID file

3, hadoop:

(1) Do not include data files in the copy, which may lead to duplicate UUID and versionID, resulting in inconsistent DN status. It is better to create a data folder

(2) Modify the project corresponding to HDFS-site. XML(3) core-site. XML modify ha.zookeeper.quorum

(4) Slaves in NN add new DN

4. Copy the cached files in the default directory of the Journode to the corresponding location on the new data node

5. In the new DN, start ZooKeeper first, then log nodes and data nodes

sbin/hadoop-daemon.sh start journalnode

sbin/hadoop-daemon.sh start datenode