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

Posted to dev@reef.apache.org by "Dhruv Mahajan (JIRA)" <ji...@apache.org> on 2015/06/28 23:27:04 UTC

[jira] [Closed] (REEF-418) Group Communication breaks down on actual cluster

     [ https://issues.apache.org/jira/browse/REEF-418?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Dhruv Mahajan closed REEF-418.
------------------------------

> Group Communication breaks down on actual cluster
> -------------------------------------------------
>
>                 Key: REEF-418
>                 URL: https://issues.apache.org/jira/browse/REEF-418
>             Project: REEF
>          Issue Type: Bug
>          Components: REEF.NET
>            Reporter: Dhruv Mahajan
>            Assignee: Dhruv Mahajan
>             Fix For: 0.12
>
>
> After the recent merge (see REEF-260), we found that once we run PipelineBroadcastAndBroadcast Reduce on actual cluster it fails while if we run it locally it passes. We found the issue to be in the WritableNetworkService where while registering observer we map it to a particular local  IP address rather than any. 
> The aim of this JIRA is to make this one line change in the WritableNetworkService.



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