[
https://issues.apache.org/jira/browse/FLINK-17072?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
ASF GitHub Bot updated FLINK-17072:
-----------------------------------
Labels: pull-request-available (was: )
> Let Dispatcher and ResourceManager pick random endpoint id
> ----------------------------------------------------------
>
> Key: FLINK-17072
> URL: https://issues.apache.org/jira/browse/FLINK-17072
> Project: Flink
> Issue Type: Improvement
> Components: Runtime / Coordination
> Affects Versions: 1.10.0
> Reporter: Till Rohrmann
> Assignee: Till Rohrmann
> Priority: Minor
> Labels: pull-request-available
> Fix For: 1.11.0
>
>
> As it is recommended by Akka, actors should not reuse the same actor name of
> a previous instance. The problem is that one can receive messages which were
> intended for the previous instance. Hence, I propose to let the
> {{ResourceManager}} and the {{Dispatcher}} pick a random name.
> A nice side effect is that the {{TestingMiniCluster}} needs no longer to use
> specialized factories for these two components in order to avoid name clashes
> when starting stand-by RM and {{Dispatchers}}. Moreover, it simplifies the
> setup of these components.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
