[
https://issues.apache.org/jira/browse/HBASE-14163?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14647317#comment-14647317
]
Samir Ahmic commented on HBASE-14163:
-------------------------------------
bq. How long did it take for your hbase master to shutdown?
In my case 10-15s.
Here is my output from $hbase master stop:
{code}
2015-07-30 09:22:20,826 INFO [main] zookeeper.ZooKeeper: Initiating client
connection, connectString=localhost:2181 sessionTimeout=90000
watcher=hconnection-0x344ec9c40x0, quorum=localhost:2181, baseZNode=/hbase
2015-07-30 09:22:20,856 INFO [main-SendThread(localhost.localdomain:2181)]
zookeeper.ClientCnxn: Opening socket connection to server
localhost.localdomain/127.0.0.1:2181. Will not attempt to authenticate using
SASL (unknown error)
2015-07-30 09:22:20,861 INFO [main-SendThread(localhost.localdomain:2181)]
zookeeper.ClientCnxn: Socket connection established to
localhost.localdomain/127.0.0.1:2181, initiating session
2015-07-30 09:22:20,870 INFO [main-SendThread(localhost.localdomain:2181)]
zookeeper.ClientCnxn: Session establishment complete on server
localhost.localdomain/127.0.0.1:2181, sessionid = 0x14eddd660110008, negotiated
timeout = 40000
2015-07-30 09:22:21,686 INFO [main] client.ConnectionImplementation: Closing
master protocol: MasterService
2015-07-30 09:22:21,686 INFO [main] client.ConnectionImplementation: Closing
zookeeper sessionid=0x14eddd660110008
2015-07-30 09:22:21,687 INFO [main] zookeeper.ZooKeeper: Session:
0x14eddd660110008 closed
2015-07-30 09:22:21,687 INFO [main-EventThread] zookeeper.ClientCnxn:
EventThread shut down
{code}
I have seen this strange behavior on OS X before in most cases it was network
related or something with fact that in default configuration we write data in
/tmp dir.
> hbase master stop loops both processes forever
> ----------------------------------------------
>
> Key: HBASE-14163
> URL: https://issues.apache.org/jira/browse/HBASE-14163
> Project: HBase
> Issue Type: Bug
> Components: master
> Affects Versions: 2.0.0
> Reporter: Allen Wittenauer
>
> It would appear that there is an infinite loop in the zk client connection
> code when performing a master stop when no external zk servers are configured.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
