hbase-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Lars Hofhansl (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HBASE-10969) TestDistributedLogSplitting fails frequently in 0.94.
Date Tue, 15 Apr 2014 04:10:14 GMT

    [ https://issues.apache.org/jira/browse/HBASE-10969?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13969196#comment-13969196
] 

Lars Hofhansl commented on HBASE-10969:
---------------------------------------

This time the test timed out. And I see this:
{code}
2014-04-14 06:28:08,502 WARN  [pool-1-thread-1] util.MBeans(73): Hadoop:service=DataNode,name=FSDatasetState-UndefinedStorageId138916243
javax.management.InstanceNotFoundException: Hadoop:service=DataNode,name=FSDatasetState-UndefinedStorageId138916243
	at com.sun.jmx.interceptor.DefaultMBeanServerInterceptor.getMBean(DefaultMBeanServerInterceptor.java:1094)
	at com.sun.jmx.interceptor.DefaultMBeanServerInterceptor.exclusiveUnregisterMBean(DefaultMBeanServerInterceptor.java:415)
	at com.sun.jmx.interceptor.DefaultMBeanServerInterceptor.unregisterMBean(DefaultMBeanServerInterceptor.java:403)
	at com.sun.jmx.mbeanserver.JmxMBeanServer.unregisterMBean(JmxMBeanServer.java:507)
	at org.apache.hadoop.metrics2.util.MBeans.unregister(MBeans.java:71)
	at org.apache.hadoop.hdfs.server.datanode.FSDataset.shutdown(FSDataset.java:2067)
	at org.apache.hadoop.hdfs.server.datanode.DataNode.shutdown(DataNode.java:799)
	at org.apache.hadoop.hdfs.MiniDFSCluster.shutdownDataNodes(MiniDFSCluster.java:566)
	at org.apache.hadoop.hdfs.MiniDFSCluster.shutdown(MiniDFSCluster.java:550)
	at org.apache.hadoop.hbase.HBaseTestingUtility.shutdownMiniDFSCluster(HBaseTestingUtility.java:475)
	at org.apache.hadoop.hbase.HBaseTestingUtility.shutdownMiniCluster(HBaseTestingUtility.java:716)
	at org.apache.hadoop.hbase.master.TestDistributedLogSplitting.after(TestDistributedLogSplitting.java:130)
	at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
{code}

So looks like I traded a race for stopping/starting the HBase mini cluster with a race further
down the stack.
The test in trunk actually starts dfs/zk only once. I port that part back as well, that would
avoid this race and also speed up the test.
Addendum coming soon.

> TestDistributedLogSplitting fails frequently in 0.94.
> -----------------------------------------------------
>
>                 Key: HBASE-10969
>                 URL: https://issues.apache.org/jira/browse/HBASE-10969
>             Project: HBase
>          Issue Type: Bug
>          Components: test
>            Reporter: Lars Hofhansl
>            Assignee: Lars Hofhansl
>             Fix For: 0.94.19
>
>         Attachments: 10969.txt
>
>
> Example:
> {code}
> Error Message
> null not an instance of org.apache.hadoop.hbase.MiniHBaseCluster
> Stacktrace
> java.lang.RuntimeException: null not an instance of org.apache.hadoop.hbase.MiniHBaseCluster
> 	at org.apache.hadoop.hbase.HBaseTestingUtility.getMiniHBaseCluster(HBaseTestingUtility.java:701)
> 	at org.apache.hadoop.hbase.HBaseTestingUtility.getHBaseCluster(HBaseTestingUtility.java:1653)
> 	at org.apache.hadoop.hbase.master.TestDistributedLogSplitting.after(TestDistributedLogSplitting.java:125)
> {code}



--
This message was sent by Atlassian JIRA
(v6.2#6252)

Mime
View raw message