[
https://issues.apache.org/jira/browse/FLINK-17196?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17086485#comment-17086485
]
Yang Wang commented on FLINK-17196:
-----------------------------------
Just one quick reminder, if we completely separate the service exposed type and
do not have any fallback, we should make sure that it could be easily used both
in managed and unmanned K8s cluster.
Since the K8s may need some time to create the load balancer, the Flink client
may take longer time for the submission. When the creation of load balancer
timeouts, we need to show up some logs for users to set the service exposed
type to {{NodePort}} or check why the LB could not be allocated.
> Improve the implementation of Fabric8FlinkKubeClient#getRestEndpoint
> --------------------------------------------------------------------
>
> Key: FLINK-17196
> URL: https://issues.apache.org/jira/browse/FLINK-17196
> Project: Flink
> Issue Type: Improvement
> Components: Deployment / Kubernetes
> Affects Versions: 1.10.0, 1.10.1
> Reporter: Canbin Zheng
> Priority: Major
> Fix For: 1.11.0
>
>
> Currently there are some bugs in the
> {{Fabric8FlinkKubeClient#getRestEndpoint}} and serveral implicit
> fallback/toleration behaviors when retrieving the Endpoint.
> This ticket proposes to fix the bugs and improve the implementation by
> deprecating some implicit fallback/toleration behaviors.
>
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
