[
https://issues.apache.org/jira/browse/FLINK-13964?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16923244#comment-16923244
]
TisonKun commented on FLINK-13964:
----------------------------------
Technically the proposal is valid and there isn't too much effort to implement
it.
I have just thought of whether we can access {{restAddressURI}} directly in
MiniCluster scenario and specially handle this case due to its special natural.
But let's say we are likely start multiple dispatcher rm components in
MiniCluster scenario(for test mostly) it's more naturally we use leader
retrieval services.
I don't see any further concerns includes rest endpoint retriever in
cluster-side high-availability services. So +1 for the proposal.
For the (conceptually and implementation) separation of leader election
services and leader retrieval services(a.k.a name services), I am drafting a
JIRA based on a failure case I have met. It won't conflict with your proposal
that cluster-side high-availability has all access. So again, let's do as you
proposed.
> Remove usage of deprecated methods from MiniCluster
> ---------------------------------------------------
>
> Key: FLINK-13964
> URL: https://issues.apache.org/jira/browse/FLINK-13964
> Project: Flink
> Issue Type: Improvement
> Components: Runtime / Coordination
> Affects Versions: 1.10.0
> Reporter: Till Rohrmann
> Assignee: Till Rohrmann
> Priority: Major
> Fix For: 1.10.0
>
>
> With FLINK-13750 we deprecated
> {{HighAvailabilityServices#getWebMonitorRetrieverService}}. This method is
> still actively used by the {{MiniCluster}}. We should remove the usage in
> order to also support custom {{HighAvailabilityService}} implementations
> which no longer implement
> {{HighAvailabilityServices#getWebMonitorRetrieverService}}.
--
This message was sent by Atlassian Jira
(v8.3.2#803003)
