The relative logs have been cleared up in order to re-launch the HBASE, I'll collect the logs when the HBASE fail.
-----Original Message----- From: Ted Yu [mailto:yuzhih...@gmail.com] Sent: Friday, January 06, 2017 10:18 AM To: user@hbase.apache.org Subject: Re: Is the hbase root directory deleted when restart Hbase everytime Can you provide a bit more detail ? version of hbase (hadoop) tail of master log when the auto-stop happened (use pastebin - attachment may not go through) Cheers On Thu, Jan 5, 2017 at 6:02 PM, QI Congyun <congyun...@alcatel-sbell.com.cn> wrote: > Hi, > > Here is a micro full distributed HBase running subsystem, total 3 > virtual hosts are involved. One is master server hosting namenode, > resource manager and Hmaster, the other hosts are datanode, nodemanager, and > Hregionserver. > The HBASE root directory is define at "hdfs://master:8020/hbase". > > I'm very bewildered that when restart the Hbase every time, the HBASE > root directory must be cleared and deleted, otherwise the HMaster > server would stopped automatically even though re-start the HBase once again. > > I wonder if my HBASE configuration has something wrong. > > Thanks. > >
