You are viewing a plain text version of this content. The canonical link for it is here.

Posted to issues@hive.apache.org by "Andrew Sherman (JIRA)" <ji...@apache.org> on 2017/12/01 17:05:00 UTC

[jira] [Commented] (HIVE-18193) Migrate existing ACID tables to use write id per table rather than global transaction id

    [ https://issues.apache.org/jira/browse/HIVE-18193?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16274638#comment-16274638 ] 

Andrew Sherman commented on HIVE-18193:
---------------------------------------

Hi [~anishek] can you explain a little more about this. How is the write id generated? Will this require on-disk data to be rewritten?
Sorry if I missed previous discussions but is there a design for snapshot isolation. Thanks

> Migrate existing ACID tables to use write id per table rather than global transaction id
> ----------------------------------------------------------------------------------------
>
>                 Key: HIVE-18193
>                 URL: https://issues.apache.org/jira/browse/HIVE-18193
>             Project: Hive
>          Issue Type: Bug
>          Components: HiveServer2, Transactions
>    Affects Versions: 3.0.0
>            Reporter: anishek
>            Assignee: anishek
>             Fix For: 3.0.0
>
>
> dependent upon HIVE-18192 
> For existing ACID Tables we need to update the table level write id metatables/sequences so any new operations on these tables works seamlessly without any conflicting data in existing base/delta files. 



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)