reef-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Tae-Geon Um (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (REEF-1661) RejectedExecutionException thrown when closing the acceptor in NettyMessageTransport
Date Fri, 17 Mar 2017 06:59:41 GMT

    [ https://issues.apache.org/jira/browse/REEF-1661?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15929505#comment-15929505
] 

Tae-Geon Um commented on REEF-1661:
-----------------------------------

[~motus] This is a minor error caused by closing {{NameResolver}} twice in {{NetworkConnectionServiceTest}}.
I've fixed the test code and create a [PR#1267|https://github.com/apache/reef/pull/1267]

> RejectedExecutionException thrown when closing the acceptor in NettyMessageTransport
> ------------------------------------------------------------------------------------
>
>                 Key: REEF-1661
>                 URL: https://issues.apache.org/jira/browse/REEF-1661
>             Project: REEF
>          Issue Type: Sub-task
>          Components: Wake
>            Reporter: Sergiy Matusevych
>            Assignee: Sergiy Matusevych
>            Priority: Minor
>
> Sometimes when the application closes Wake {{NettyMessageTransport}}, and exception is
thrown from {{acceptor.close()}} call. We catch and log the error and then proceed with the
shutdown; this is normal behavior required to make the  {{NettyMessageTransport.close()}}
method idempotent.
> The exception does not prevent us from the graceful shutdown of the transport, but it
is still unclear what exactly causes the error. Do we close the transport more than once?
Does the acceptor gets closed as a side effect of some other service shutdown? The error might
be an indicator of some minor error or a race condition somewhere in the REEF/Wake shutdown
process.



--
This message was sent by Atlassian JIRA
(v6.3.15#6346)

Mime
View raw message