Reason: power failure (including computer sleep, etc.) caused Hmaster to fail to connect, and no master node could be found in zookeeper

Solution: delete the hbase node in zookeeper, open hbase will automatically create this node
1) log in to the zookeeper client: zkCli.sh
2) delete the hbase node: deleteall /hbase

The most critical step: restart hbase, restart zookeeper
1) close zookeeper: my_zk.sh stop This is my script to start zk, don’t copy
2) close hbase: stop-hbase.sh is invalid, use the jps command to find each in the cluster The port number of the hmaster and hregionserver of the machine, kill -9 + port number one by one kills the hbase process, which is equivalent to manually closing hbase
3) Open zookeeper and hbase, my_zk.sh start start-hbase.sh
hdfs does not need to be moved, If yours doesn’t work, hdfs can also be restarted.