You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@hive.apache.org by "ASF GitHub Bot (Jira)" <ji...@apache.org> on 2023/01/16 09:15:00 UTC

[jira] [Work logged] (HIVE-26597) Fix unsetting of db prop repl.target.for in ReplicationSemanticAnalyzer

     [ https://issues.apache.org/jira/browse/HIVE-26597?focusedWorklogId=839336&page=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-839336 ]

ASF GitHub Bot logged work on HIVE-26597:
-----------------------------------------

                Author: ASF GitHub Bot
            Created on: 16/Jan/23 09:14
            Start Date: 16/Jan/23 09:14
    Worklog Time Spent: 10m 
      Work Description: pudidic commented on PR #3788:
URL: https://github.com/apache/hive/pull/3788#issuecomment-1383713734

   Looks good to me, but orc_llap_counters.q is failing. Please push an empty commit with `git commit --allow-empty -m 'Re-trigger CI'`.




Issue Time Tracking
-------------------

    Worklog Id:     (was: 839336)
    Time Spent: 0.5h  (was: 20m)

> Fix unsetting of db prop repl.target.for in ReplicationSemanticAnalyzer
> -----------------------------------------------------------------------
>
>                 Key: HIVE-26597
>                 URL: https://issues.apache.org/jira/browse/HIVE-26597
>             Project: Hive
>          Issue Type: Bug
>            Reporter: Teddy Choi
>            Assignee: Rakshith C
>            Priority: Major
>              Labels: pull-request-available
>          Time Spent: 0.5h
>  Remaining Estimate: 0h
>
> when repl policy is set from A -> B
>  * *repl.target.for* is set on B.
> when failover is initiated
>  * *repl.failover.endpoint* = *'TARGET'* is set on B.
>  
> now when reverse policy is set up from {*}A <- B{*};
> there is a check in [ReplicationSemanticAnalyzer#initReplDump|https://github.com/apache/hive/blob/master/ql/src/java/org/apache/hadoop/hive/ql/parse/ReplicationSemanticAnalyzer.java#L196] which checks for existence of these two properties and if they are set,
> it unsets the *repl.target.for* property.
> Because of this optimisedBootstrap won't be triggered because it checks for the existence of *repl.target.for* property during repl dump on target [HERE|https://github.com/apache/hive/blob/master/ql/src/java/org/apache/hadoop/hive/ql/exec/repl/OptimisedBootstrapUtils.java#L93].
>  
> Fix : remove the code which unsets repl.target.for in ReplicationSemanticAnalyzer, because second dump cycle of optimized bootstrap unsets it.



--
This message was sent by Atlassian Jira
(v8.20.10#820010)