reef-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Andrew Chung (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (REEF-1279) Injecting RuntimeClock in event handler creates second instance of clock
Date Thu, 24 Mar 2016 23:35:25 GMT

    [ https://issues.apache.org/jira/browse/REEF-1279?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15211135#comment-15211135
] 

Andrew Chung commented on REEF-1279:
------------------------------------

[~MariiaMykhailova] I've been checking it out and it might not actually be as hard as I thought
it would. Will provide an update. Thanks!

> Injecting RuntimeClock in event handler creates second instance of clock
> ------------------------------------------------------------------------
>
>                 Key: REEF-1279
>                 URL: https://issues.apache.org/jira/browse/REEF-1279
>             Project: REEF
>          Issue Type: Bug
>          Components: REEF.NET
>            Reporter: Mariia Mykhailova
>            Assignee: Andrew Chung
>
> For REEF-1249 I've created an event handler which uses {{RuntimeClock}} to trigger an
exception some time after the event. {{RuntimeClock}}  is injected, so it should be the same
instance as the one used in Evaluator, and exception thrown on alarm should cause some failure.
> However, this is not the case. Second instance of {{RuntimeClock}} is created (so it
requires a separate call to clock.Run(), starts a clock in separate thread, and the thrown
exception is ignored).
> My code with a repro can be found at https://github.com/tcNickolas/reef/tree/Second-Clock-Repro



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

Mime
View raw message