[jira] [Updated] (AMBARI-19834) HDP 3.0 support for Spark with configs, kerberos, widgets, metrics, quicklinks, and themes

["Dmytro Sen (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/AMBARI-19834?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Dmytro Sen updated AMBARI-19834:
--------------------------------
    Attachment: AMBARI-19834.patch

> HDP 3.0 support for Spark with configs, kerberos, widgets, metrics, quicklinks, and themes
> ------------------------------------------------------------------------------------------
>
>                 Key: AMBARI-19834
>                 URL: https://issues.apache.org/jira/browse/AMBARI-19834
>             Project: Ambari
>          Issue Type: Story
>          Components: stacks
>    Affects Versions: 3.0.0
>            Reporter: Alejandro Fernandez
>             Fix For: 3.0.0
>
>         Attachments: AMBARI-19834.patch
>
>
> HDP 3.0 support for Spark with configs, kerberos, widgets, metrics, quicklinks, and themes
> Flatten from HDP 2.0.6 - 2.6 into common-services, and reference in HDP 3.0
> IMPORTANT: In HDP 3.0, Spark1 is being removed, and Spark2 is being rebranded as "Spark"
> In HDP 3.0, we have created a new stack definition that does not inherit from other stacks, in order to reduce the complexity of having to analyze older stacks.
> This means that we need to create a service definition (metainfo.xml, configs, kerberos, widgets, metrics, quicklinks, and themes) that is equivalent to what is inherit and deleted from all of the previous stacks.
> A merge needs to account for additions, overrides, and deletions.
> metainfo.xml and configs perform a merge of older versions
> kerberos.json always seems to override the previous file
> Because the bits for this service may not yet be available in the HDP 3.0 repo, 
> the task is to ensure that /api/v1/stacks/HDP/versions/2.6/services/SPARK2 (which uses inheritance) is equivalent to the flattening of /api/v1/stacks/HDP/versions/3.0/services/SPARK .
> Please take a look at how this was done for ZK, HDFS, and YARN/MR.
> This means that you will not be able to actually install the service for now, but can still perform validation during the Install Wizard that the correct components and configs show up.



--
This message was sent by Atlassian JIRA
(v6.3.15#6346)