[
https://issues.apache.org/jira/browse/HADOOP-2662?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12560632#action_12560632
]
stack commented on HADOOP-2662:
-------------------------------
So, things are working 'properly'. Master waits on each region to either report
its closed or until the regionserver's lease times out. When either has
happened for all servers, then it goes down. On regionserver side, it results
in ugly exception because its trying to talk to a master long gone. Its
retrying to connect? In this shutdown case it probably shouldn't run the
retries. It should just exit on the first failure-to-connect.. Let this issue
be about disabling retries in this shutdown case and possibly make the
exception thrown a little sweeter (though its pretty spot-on as is).
> master stopping before regions finish stopping
> ----------------------------------------------
>
> Key: HADOOP-2662
> URL: https://issues.apache.org/jira/browse/HADOOP-2662
> Project: Hadoop
> Issue Type: Bug
> Components: contrib/hbase
> Reporter: Billy Pearson
> Priority: Minor
> Fix For: 0.16.1
>
>
> I get this on my region servers logs sometimes when I shutdown the cluster it
> repeats several times sometimes trying to find the master
> I thank we need to look at the master and make sure we do not stop the master
> on exit before all region servers report down.
> I am not sure if there could be data loss or not but we should not leave
> region servers looking for the master unless it has failed on it own.
> {code}
> 2008-01-18 17:40:42,009 WARN org.apache.hadoop.hbase.HRegionServer: Failed to
> send exiting message to master:
> java.net.ConnectException: Connection refused
> at java.net.PlainSocketImpl.socketConnect(Native Method)
> at java.net.PlainSocketImpl.doConnect(PlainSocketImpl.java:333)
> at java.net.PlainSocketImpl.connectToAddress(PlainSocketImpl.java:195)
> at java.net.PlainSocketImpl.connect(PlainSocketImpl.java:182)
> at java.net.SocksSocketImpl.connect(SocksSocketImpl.java:366)
> at java.net.Socket.connect(Socket.java:519)
> at
> org.apache.hadoop.ipc.Client$Connection.setupIOstreams(Client.java:159)
> at org.apache.hadoop.ipc.Client.getConnection(Client.java:575)
> at org.apache.hadoop.ipc.Client.call(Client.java:498)
> at
> org.apache.hadoop.hbase.ipc.HbaseRPC$Invoker.invoke(HbaseRPC.java:210)
> at $Proxy0.regionServerReport(Unknown Source)
> at org.apache.hadoop.hbase.HRegionServer.run(HRegionServer.java:898)
> at java.lang.Thread.run(Thread.java:595)
> {code}
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
