[
https://issues.apache.org/jira/browse/FLINK-4451?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15433180#comment-15433180
]
ASF GitHub Bot commented on FLINK-4451:
---------------------------------------
Github user StephanEwen commented on the issue:
https://github.com/apache/flink/pull/2405
The idea makes sense. How does the system determine that the RPC endpoint
is not reachable? By the fact that the initial "identify" message times out?
> Throw exception when remote connection cannot be resolved
> ---------------------------------------------------------
>
> Key: FLINK-4451
> URL: https://issues.apache.org/jira/browse/FLINK-4451
> Project: Flink
> Issue Type: Sub-task
> Components: Distributed Coordination
> Reporter: Till Rohrmann
> Assignee: Till Rohrmann
>
> The {{RpcService}} implementation should throw an exception (returned in the
> future) if {{RpcService.connect(address, type)}} cannot connect to the remote
> {{RpcEndpoint}}.
> At the moment the {{AkkaRpcService}} does not check that the
> {{IdentifyActor}} message contains a valid {{ActorRef}} and throws due to
> that a {{NullPointerException}}.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
