hadoop-hdfs-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Steve Loughran (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HDFS-3582) Hook daemon process exit for testing
Date Tue, 10 Jul 2012 08:59:35 GMT

    [ https://issues.apache.org/jira/browse/HDFS-3582?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13410143#comment-13410143

Steve Loughran commented on HDFS-3582:

You don't need to do this for testing because the security manager hook is much less invasive.
Indeed, all you need to do is 

1. stick ant.jar on the classpath
2. set the security manager to org.apache.tools.ant.util.optional.NoExitSecurityManager:


I am supportive of a single exit point and the ability to trap exits -which is is why I field
HADOOP-8096 in the first place. But to embed Hadoop properly in OSGi we need this across the
* the exit routine and exception should go into hadoop-core
* -hdfs and -mapreduce should adopt it

Having an HDFS-specific variant purely for testing doesn't aid OSGi integration, and will
only increase the effort of switching to a unified version downstream. 

Why not fix it properly the first time, and apply it in HDFS first?
> Hook daemon process exit for testing 
> -------------------------------------
>                 Key: HDFS-3582
>                 URL: https://issues.apache.org/jira/browse/HDFS-3582
>             Project: Hadoop HDFS
>          Issue Type: Improvement
>          Components: test
>    Affects Versions: 2.0.0-alpha
>            Reporter: Eli Collins
>            Assignee: Eli Collins
>            Priority: Minor
>         Attachments: hdfs-3582.txt, hdfs-3582.txt, hdfs-3582.txt, hdfs-3582.txt, hdfs-3582.txt
> Occasionally the tests fail with "java.util.concurrent.ExecutionException: org.apache.maven.surefire.booter.SurefireBooterForkException:
> Error occurred in starting fork, check output in log" because the NN is exit'ing (via
System#exit or Runtime#exit). Unfortunately Surefire doesn't retain the log output (see SUREFIRE-871)
so the test log is empty, we don't know which part of the test triggered which exit in HDFS.
To make this easier to debug let's hook all daemon process exits when running the tests.

This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira


View raw message