Hi, Long GC pause delays the heartbeat from the region server to zookeeper, and make the connection between the region server and zookeeper timeout. Zookeeper deletes the znode of this region server after the connection is considered as expired, master receives the event and reassign regions owned by this region server. When this region server is back to work it receives the expired exception from the zookeeper, then it is aborted accordingly.
You can tune your region server to reduce the GC pause, or enlarge the zookeeper timeout configuration in hbase-site.xml (zookeeper.session.timeout) which has side-affect that it takes more time to detect the failed region server by master. Regards, Jingcheng -----Original Message----- From: 邸星星 [mailto:[email protected]] Sent: Tuesday, August 30, 2016 9:20 AM To: [email protected] Subject: RegionServer shutdown by some unknown reason. Hi : In our hbase cluster, one regionserver looked shutdown by it's self, I made an issue, can somebody help me about this ? https://issues.apache.org/jira/browse/HBASE-16514 A lot of thanks for all!
