You are viewing a plain text version of this content. The canonical link for it is here.
Posted to mapreduce-issues@hadoop.apache.org by "Niranjan Singh (JIRA)" <ji...@apache.org> on 2013/06/05 01:56:20 UTC

[jira] [Commented] (MAPREDUCE-5185) When log aggregation not enabled, message should point to NM HTTP port, not IPC port

    [ https://issues.apache.org/jira/browse/MAPREDUCE-5185?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13675448#comment-13675448 ] 

Niranjan Singh commented on MAPREDUCE-5185:
-------------------------------------------

@Sandy Ryza: I tried to reproduce the scenario without log aggregation enabled,but was unable to get the message "Aggregation is not enabled.....".I was looking at mapred-niranjan-historyserver-niranjan-OEM.log file. But somehow I came up with the solution. Below is the code:

if (!conf.getBoolean(YarnConfiguration.LOG_AGGREGATION_ENABLED,
        YarnConfiguration.DEFAULT_LOG_AGGREGATION_ENABLED)) {
      String nodeAddress = conf.get(YarnConfiguration.NM_WEBAPP_ADDRESS,
    	        YarnConfiguration.DEFAULT_NM_WEBAPP_ADDRESS);
    	   
      html.h1()
          ._("Aggregation is not enabled. Try the nodemanager at " + nodeAddress)
          ._();
      return;
    }
Will this work?
                
> When log aggregation not enabled, message should point to NM HTTP port, not IPC port 
> -------------------------------------------------------------------------------------
>
>                 Key: MAPREDUCE-5185
>                 URL: https://issues.apache.org/jira/browse/MAPREDUCE-5185
>             Project: Hadoop Map/Reduce
>          Issue Type: Bug
>          Components: jobhistoryserver
>    Affects Versions: 2.0.4-alpha
>            Reporter: Sandy Ryza
>            Assignee: Sandy Ryza
>
> When I try to get a container's logs in the JHS without log aggregation enabled, I get a message that looks like this:
> "Aggregation is not enabled. Try the nodemanager at sandy-ThinkPad-T530:33224"
> This could be a lot more helpful by actually pointing the URL that would show the container logs on the NM.

--
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