apex-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Vlad Rozov (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (APEXCORE-759) Operators connected using THREAD_LOCAL stream locality should enforce parallel partition scheme
Date Fri, 14 Jul 2017 01:31:00 GMT

    [ https://issues.apache.org/jira/browse/APEXCORE-759?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16086702#comment-16086702
] 

Vlad Rozov commented on APEXCORE-759:
-------------------------------------

Will all N downstream partitions be in the same thread? I don't think it is supported now
(thread local and container local afaik only support parallel partitions). The use case is
possible but I don't think that it will be widely accepted or used. I am not sure that it
needs to be supported.

> Operators connected using THREAD_LOCAL stream locality should enforce parallel partition
scheme
> -----------------------------------------------------------------------------------------------
>
>                 Key: APEXCORE-759
>                 URL: https://issues.apache.org/jira/browse/APEXCORE-759
>             Project: Apache Apex Core
>          Issue Type: Bug
>            Reporter: Vinay Bangalore Srikanth
>         Attachments: apex (1).log, apex.log
>
>
> In my application - http://node0.morado.com:9090/static/#/ops/apps/application_1499808956620_0190
, I have set the locality to be THREAD_LOCAL.
> Upstream operator - Random string generator (with 4 partitions)
> Downstream operator - Custom console operator (with 5 partitions)
> The containers are getting killed. Exceptions from container-launch are seen.
> Logical message has to be displayed by handling exceptions.
> Logs are attached for one of the containers - Id: 01_000126  that was killed. 



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)

Mime
View raw message