You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@oozie.apache.org by "Robert Kanter (JIRA)" <ji...@apache.org> on 2013/02/15 06:05:12 UTC

[jira] [Created] (OOZIE-1227) In a coordinator, specifying the without a namenode causes it to fail

Robert Kanter created OOZIE-1227:
------------------------------------

             Summary: In a coordinator, specifying the <app-path> without a namenode causes it to fail
                 Key: OOZIE-1227
                 URL: https://issues.apache.org/jira/browse/OOZIE-1227
             Project: Oozie
          Issue Type: Bug
    Affects Versions: 3.3.0
            Reporter: Robert Kanter
            Assignee: Robert Kanter
             Fix For: trunk, 3.3.2


In a coordinator, some users don't specify the name node in the path to the workflow application.  e.g.
{code}
<action>
    <workflow>
        <app-path>/user/${coord:user()}/${examplesRoot}/apps/aggregator</app-path>
        ...
{code}
This now causes the coordinator to fail.  I think this may also affect other paths because many of them go through the HadoopAccessorService.  

Ideally, we should fix this behavior to the old way: have it default to HDFS and get the namenode automatically if no scheme://host:port is given.  

A workaround is to to add the following property to oozie-site.xml
{code}
<property>
    <name>oozie.service.HadoopAccessorService.supported.filesystems</name>
    <value>*</value>
</property>
{code}
This causes Oozie to effectively have the old behavior.  The downside of this workaround is that administrators cannot restrict the filesystems that Oozie will allow ({{*}} means all filesystems, even non-real ones).  

Instead of making users use that workaround, we can simply make Oozie skip the check if the scheme is {{null}}.  It would allow the old behavior without disabling the scheme check.  

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira