hive-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Siddharth Seth (JIRA)" <j...@apache.org>
Subject [jira] [Updated] (HIVE-14023) LLAP: Make the Hive query id available in ContainerRunner
Date Thu, 16 Jun 2016 20:21:05 GMT

     [ https://issues.apache.org/jira/browse/HIVE-14023?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]

Siddharth Seth updated HIVE-14023:
----------------------------------
    Attachment: HIVE-14023.01.patch

Patch to propagate the queryId to LLAP.

While testing this, I noticed that the same queryId was used across multiple Dags. This happened
in a couple of case. This was while running TestMiniLlapCliDriver. Will try again to see what's
going on and create a jira. Thought the queryId was supposed to be unique.

[~sershe] - please review.
Also, to investigate queryId not being unique, do you know whether the query string is accessible
from within the TezProcessor (from the conf) ?

> LLAP: Make the Hive query id available in ContainerRunner
> ---------------------------------------------------------
>
>                 Key: HIVE-14023
>                 URL: https://issues.apache.org/jira/browse/HIVE-14023
>             Project: Hive
>          Issue Type: Improvement
>            Reporter: Siddharth Seth
>            Assignee: Siddharth Seth
>         Attachments: HIVE-14023.01.patch
>
>
> Needed to generate logs per query.
> We can use the dag identifier for now, but that isn't very useful. (The queryId may not
be too useful either if users cannot find it - but that's better than a dagIdentifier)
> The queryId is available right now after the Processor starts, which is too late for
log changes.



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

Mime
View raw message