You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@flink.apache.org by "Leonard Xu (Jira)" <ji...@apache.org> on 2020/12/01 14:39:00 UTC

[jira] [Commented] (FLINK-20341) Rework Deployment / Coordination Documentation

    [ https://issues.apache.org/jira/browse/FLINK-20341?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17241579#comment-17241579 ] 

Leonard Xu commented on FLINK-20341:
------------------------------------

Hi, [~rmetzger] [~trohrmann]


Kindly remind,  we'd better use \{% link deployment/config.zh.md %}  rather than \{% link deployment/config.md %} in *.zh.md doc, otherwise the document build(run comand flink/docs/build_docs.sh) will fail with following error:

 
{code:java}
Liquid Exception: Could not find document 'deployment/config.md' in tag 'link'. Make sure the document exists and the path is correct. in deployment/ha/zookeeper_ha.zh.md
  Liquid Exception: Could not find document 'deployment/index.md' in tag 'link'. Make sure the document exists and the path is correct. in deployment/resource-providers/yarn.zh.md
...{code}
 

I've saw several failed pages that merged recently and comes from the subtask of this umbrella.

I think we need add a subtask to fix these pages ?

 

> Rework Deployment / Coordination Documentation
> ----------------------------------------------
>
>                 Key: FLINK-20341
>                 URL: https://issues.apache.org/jira/browse/FLINK-20341
>             Project: Flink
>          Issue Type: Task
>          Components: Documentation, Runtime / Coordination
>            Reporter: Robert Metzger
>            Assignee: Robert Metzger
>            Priority: Major
>             Fix For: 1.12.0
>
>
> Problems:
> - Clusters & Deployment pages are very inhomogeneous
> - Overview page has good intentions, but is a huge wall of text
>  - Native K8s and YARN have a “Background / Internals” page
>  - difference between Local Cluster and Standalone Cluster is unclear
> Goals:
> - Deploying a Flink cluster is one of the first tasks when getting to know Flink. We need proper guidance for making these steps a success
> - We need a proper separation between general concepts (HA, session/perjob mode) and implementations of them (ZK HA, K8s HA, YARN session, …). Also orthogonal aspects such as FileSystems, Plugins, Security etc.
> Related work:
> https://cwiki.apache.org/confluence/display/FLINK/FLIP-42%3A+Rework+Flink+Documentation (see “Deployment Section”)



--
This message was sent by Atlassian Jira
(v8.3.4#803005)