[
https://issues.apache.org/jira/browse/FLINK-21382?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17297904#comment-17297904
]
Peng Zhang commented on FLINK-21382:
------------------------------------
I also met the same problem:
* tried to have 2 JobManagers and got error, and not sure what is wrong (as
said in this issue, only 1 JobManager could work). Then, question is that is it
possible to set up 2 JobManagers (I suppose this is what HA means)?
* use one JobManager and HA, but also got other issues
https://issues.apache.org/jira/browse/FLINK-21685 and
https://issues.apache.org/jira/browse/FLINK-21472
So, I am not able yet to run successfully Flink HA on K8S (1 JobManager and a
few TaskManagers in session cluster mode)
I have not tried Flink on native k8s, but there seems to me not enough
documents either (e.g., I would prefer to use configmaps for the parameters
than passing properties in command line since we may need to tune a number
flink configurations for our jobs, especially memory configurations. It is not
clear to me in native k8s, could I use configmaps or not?) Thanks!
> Standalone K8s documentation does not explain usage of standby JobManagers
> --------------------------------------------------------------------------
>
> Key: FLINK-21382
> URL: https://issues.apache.org/jira/browse/FLINK-21382
> Project: Flink
> Issue Type: Improvement
> Components: Deployment / Kubernetes, Documentation
> Affects Versions: 1.12.1, 1.13.0
> Reporter: Till Rohrmann
> Assignee: Yang Wang
> Priority: Major
>
> Our [standalone K8s
> documentation|https://ci.apache.org/projects/flink/flink-docs-release-1.12/deployment/resource-providers/standalone/kubernetes.html#high-availability-with-standalone-kubernetes]
> mentions how to configure K8s HA services. It does not mention that this
> only works with a single JobManager. When using standby JobManagers, then the
> given deployment yamls won't work because the {{jobmanager.rpc.address}} is
> configured to be the {{jobmanager}} service.
> Changing the configuration to work is surprisingly difficult because of a
> lack of documentation. Moreover, it is quite difficult to pass in custom
> configuration values when using a ConfigMap for sharing Flink's
> {{flink-conf.yaml}}. The problem is that mounted ConfigMaps are not writable
> from a pod perspective. See [this
> answer|https://stackoverflow.com/a/66228073/4815083] for how one could
> achieve it.
> I think we could improve our documentation to explain our users how to
> configure a standalone HA cluster with standby JobManagers.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
