[jira] [Work logged] (HIVE-27107) Fix improper metrics count for RESUME/RESET workflow

["ASF GitHub Bot (Jira)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/HIVE-27107?focusedWorklogId=847735&page=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-847735 ]

ASF GitHub Bot logged work on HIVE-27107:
-----------------------------------------

                Author: ASF GitHub Bot
            Created on: 27/Feb/23 07:43
            Start Date: 27/Feb/23 07:43
    Worklog Time Spent: 10m 
      Work Description: shreenidhiSaigaonkar opened a new pull request, #4084:
URL: https://github.com/apache/hive/pull/4084

   
   
   ### What changes were proposed in this pull request?
   This patch fixes the improper metrics (TABLE count) incase of RESUME/RESET by taking intersection of tables in ```table_diff_complete``` and actual ```tables on source```. 
   
   
   
   ### Why are the changes needed?
   ```REPL DUMP``` during RESET checks all the tables listed in ```table_diff_complete``` and persists
   the table count in replication metrics. This is incorrect for RESET,
   because table_diff_complete has those tables which needs to be dropped too.
   
   
   
   ### Does this PR introduce _any_ user-facing change?
   NO
   
   
   
   ### How was this patch tested?
   * Updated unit test for this scenario.
   * Tested it in local setup.
   
   




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

            Worklog Id:     (was: 847735)
    Remaining Estimate: 0h
            Time Spent: 10m

> Fix improper metrics count for RESUME/RESET workflow
> ----------------------------------------------------
>
>                 Key: HIVE-27107
>                 URL: https://issues.apache.org/jira/browse/HIVE-27107
>             Project: Hive
>          Issue Type: Task
>            Reporter: Shreenidhi
>            Assignee: Shreenidhi
>            Priority: Major
>          Time Spent: 10m
>  Remaining Estimate: 0h
>
> *Analysis :* 
> REPL DUMP during RESET checks all the tables listed in table_diff_complete
> and persists the table count in replication metrics. This is incorrect for RESET, 
> because table_diff_complete has those tables which needs to be dropped too.
> *Fix :* 
> To fix this, one can take intersection of (tables in table_diff_complete, actual tables on source) and capture the same into the replication_metrics



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