You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@phoenix.apache.org by "Josh Elser (JIRA)" <ji...@apache.org> on 2018/07/09 19:29:00 UTC

[jira] [Commented] (PHOENIX-4806) make PhoenixStorageHander using only unmanaged tables

    [ https://issues.apache.org/jira/browse/PHOENIX-4806?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16537440#comment-16537440 ] 

Josh Elser commented on PHOENIX-4806:
-------------------------------------

[~sergey.soldatov], do you know if this change in semantics landed into the Hive-3.1.0 release? Or, more generally, when we'll see it in a Hive release?

> make PhoenixStorageHander using only unmanaged tables
> -----------------------------------------------------
>
>                 Key: PHOENIX-4806
>                 URL: https://issues.apache.org/jira/browse/PHOENIX-4806
>             Project: Phoenix
>          Issue Type: Bug
>    Affects Versions: 4.14.0, 5.0.0
>            Reporter: Sergey Soldatov
>            Assignee: Sergey Soldatov
>            Priority: Major
>
> For managed table Hive is heavily using internal statistic which is not updated properly when data is changing from outside. So all custom storage handlers are moving to the schema when only unmanaged (external) tables are using.
> The suggested lifecycle:
> 1) All tables should be created with external keyword
> 2) if Phoenix table doesn't exist, we create a new one and add a flag that it should be deleted on drop table from hive
> 3) If Phoenix table exists, no changes in the existing behavior.
> Some of the storage handlers that are part of Hive distribution are already using this schema. Others are in progress.
> FYI [~elserj]



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)