You are viewing a plain text version of this content. The canonical link for it is here.

Posted to issues@hive.apache.org by "Prasanth Jayachandran (JIRA)" <ji...@apache.org> on 2016/04/24 07:50:12 UTC

[jira] [Resolved] (HIVE-13600) LLAP: Web service selector channel thread is getting completely blocked

     [ https://issues.apache.org/jira/browse/HIVE-13600?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Prasanth Jayachandran resolved HIVE-13600.
------------------------------------------
    Resolution: Invalid

Actually this is invalid. After further investigation it looks like most of the tasks are blocked on synchronous logging. 

> LLAP: Web service selector channel thread is getting completely blocked
> -----------------------------------------------------------------------
>
>                 Key: HIVE-13600
>                 URL: https://issues.apache.org/jira/browse/HIVE-13600
>             Project: Hive
>          Issue Type: Bug
>    Affects Versions: 2.1.0
>            Reporter: Prasanth Jayachandran
>         Attachments: Screen Shot 2016-04-24 at 12.27.55 AM.png
>
>
> With pre-emption enabled, 4 executors and with wait queue size of 8. Repeatedly running tcpds queries 27 and 5 on scale 2 resulted in deadlock. 
> Not sure about the cause yet or if this is the real cause. Yourkit profile shows llap web selector channel thread being blocked. See attached image.



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