[ 
https://issues.apache.org/jira/browse/HIVE-3215?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13403587#comment-13403587
 ] 

Zhenxiao Luo commented on HIVE-3215:
------------------------------------

@Ramkumar:

Generally good to me, two things:

#1. I think it is OK to keep the following:
-        String taskUrl = (jtUrl == null) ? "Unavailable" :
-        jtUrl + "/taskdetails.jsp?jobid=" + jobId + "&tipid=" + 
task.toString();

-          sb.append("URL:\n  " + taskUrl + "\n");

Is there a reason to get rid of these info?

#2. I just checked, the only usage of JobTrackerURLResolver is:

[~/Code/studyHive]find . -name *.java | xargs grep JobTrackerURLResolver.
./ql/src/java/org/apache/hadoop/hive/ql/exec/JobDebugger.java:      jtUrl = 
JobTrackerURLResolver.getURL(conf);
./ql/src/java/org/apache/hadoop/hive/ql/exec/JobTrackerURLResolver.java: * 
JobTrackerURLResolver.
./ql/src/java/org/apache/hadoop/hive/ql/exec/JobTrackerURLResolver.java:public 
final class JobTrackerURLResolver {
./ql/src/java/org/apache/hadoop/hive/ql/exec/JobTrackerURLResolver.java:  
private JobTrackerURLResolver() {
./ql/src/java/org/apache/hadoop/hive/ql/exec/HadoopJobExecHelper.java:    
String jtUrl = JobTrackerURLResolver.getURL(conf);
./ql/src/java/org/apache/hadoop/hive/ql/exec/Throttle.java:      String tracker 
= JobTrackerURLResolver.getURL(conf) + "/gc.jsp?threshold=" + threshold;

How about also update Throttle.java, and retire JobTrackerURLResolver.java at 
all?

Thanks,
Zhenxiao
                
> JobDebugger should use RunningJob.getTrackingURL 
> -------------------------------------------------
>
>                 Key: HIVE-3215
>                 URL: https://issues.apache.org/jira/browse/HIVE-3215
>             Project: Hive
>          Issue Type: Bug
>          Components: CLI
>            Reporter: Ramkumar Vadali
>            Assignee: Ramkumar Vadali
>            Priority: Minor
>         Attachments: HIVE-3215.2.patch, HIVE-3215.patch
>
>
> When a MR job fails, the JobDebugger tries to construct the job tracker URL 
> by connecting to the job tracker, but that is better done by using 
> RunningJob#getTrackingURL.
> Also, it tries to construct URLs to the tasks, which is not reliable, because 
> the job could have been retired and the URL would not work.

--
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

        

Reply via email to