[ 
https://issues.apache.org/jira/browse/MAPREDUCE-6565?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15290467#comment-15290467
 ] 

Yuren Wu commented on MAPREDUCE-6565:
-------------------------------------

further thoughts on this fix.. the multi-home network setup for hadoop is 
getting some attention in industry.  The security token design using toke + 
service name does not have updates to accommodate the complex network setup. HA 
This quick fix just get by with executing mapreduce jobs. However, I would 
suggest to create a new request to address the multi-home network and token 
handling in a more organized effort. Security package has very succinct log and 
it took me quite while to track down the issue. Properties under 
hadoop.security should be handled in a single code base to interact with 
various components. Credential token services such as retrieve/clone are 
handled by different packages in multiple components. Current code is really 
difficult to understand and manage. 

> Configuration to use host name in delegation token service is not read from 
> job.xml during MapReduce job execution.
> -------------------------------------------------------------------------------------------------------------------
>
>                 Key: MAPREDUCE-6565
>                 URL: https://issues.apache.org/jira/browse/MAPREDUCE-6565
>             Project: Hadoop Map/Reduce
>          Issue Type: Bug
>            Reporter: Chris Nauroth
>
> By default, the service field of a delegation token is populated based on 
> server IP address.  Setting {{hadoop.security.token.service.use_ip}} to 
> {{false}} changes this behavior to use host name instead of IP address.  
> However, this configuration property is not read from job.xml.  Instead, it's 
> read from a separate {{Configuration}} instance created during static 
> initialization of {{SecurityUtil}}.  This does not work correctly with 
> MapReduce jobs if the framework is distributed by setting 
> {{mapreduce.application.framework.path}} and the 
> {{mapreduce.application.classpath}} is isolated to avoid reading 
> core-site.xml from the cluster nodes.  MapReduce tasks will fail to 
> authenticate to HDFS, because they'll try to find a delegation token based on 
> the NameNode IP address, even though at job submission time the tokens were 
> generated using the host name.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

---------------------------------------------------------------------
To unsubscribe, e-mail: mapreduce-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: mapreduce-issues-h...@hadoop.apache.org

Reply via email to