hadoop-hive-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Edward Capriolo (JIRA)" <j...@apache.org>
Subject [jira] Commented: (HIVE-551) Metrics for Hive
Date Fri, 24 Jul 2009 20:23:14 GMT

    [ https://issues.apache.org/jira/browse/HIVE-551?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12735146#action_12735146
] 

Edward Capriolo commented on HIVE-551:
--------------------------------------

Saw something JMX related in another Jira and wanted to bring this up

Suppose we had a configuration option like...
{noformat}
set hive.jmx.server.per.job=true
set hive.jmx.server.user=ed
set hive.jmx.server.password=ed
{noformat}

A hive query would emit a status line like...
{noformat}
jmx url = service:///jmx/hostname:port
{noformat}

Different types of queries would be free to add objects to the MBean server and then someone
could attach in and profile a job as running.

> Metrics for Hive
> ----------------
>
>                 Key: HIVE-551
>                 URL: https://issues.apache.org/jira/browse/HIVE-551
>             Project: Hadoop Hive
>          Issue Type: Wish
>          Components: Logging
>            Reporter: Edward Capriolo
>            Priority: Minor
>
> It would be great to have metrics possibly totalQueries, failedQueries, totalQueryTime,
FilesScanned, the challenge here is that multiple instanced of hive could be running, hive
server,hive cli, hivecli so having a dedicated JMX port like in the configuration file would
be difficult. The user could set the JMX port through the CLI. Any thoughts?

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


Mime
View raw message