Disable and drop cannot be executed in hbase shell

1 Querying work status through the Web monitoring page of HBase discover the regions in transition area has a long-running task, you need to close Hbase-master and restart, and then this problem process I directly use/etc/init.d/ Hbase-master restar command Cluster and perform the operation, none of them take effect, finally execute/etc/init.d/hbase-master stop and the region in the cluster The hbase-regisionserver machine for num 0 machine passes/etc/init.d/hbase-regionserver stop and checks if the program process is off, When I execute the JPS again, the phenomenon cannot be viewed directly to the associated process name corresponding to the process number.

[[email protected] ~]# jps71125 Jps143011 NodeManager65046 MRAppMaster59326 DataNode66658 -- process information unavailable66719 -- process information unavailable

If this is the case, you can use the method to view, Java every time you start a process and/tmp directory, there will be a separate record as follows:
ls /tmp/hsperfdata_*Go to the appropriate directory to see the process number, if there is a command executed above did not succeed, can be closed by kill-9 PID, and then enable the cluster

Shell>/etc/init.d/hbase-master start
Shell>hbase Shell
Hbase>balance_switch true

HBase (Main):007:0> Balance_switch true #开启hbase的balance重新平衡
False
0 row (s) in 0.0070 seconds

HBase (main):008:0> balancer #手动触发balance
True
0 row (s) in 57.5330 seconds

The view regions in transition is now normal, and the Regions.num area also restores balance
The Disable and drop commands are returned to normal when entering the hbase shell environment
Cases:

hbase(main):001:0> drop ‘hf_csql_20180301‘0 row(s) in 0.6670 seconds

Disable and drop cannot be executed in hbase shell