[ 
https://issues.apache.org/jira/browse/HIVE-26167?focusedWorklogId=760755&page=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-760755
 ]

ASF GitHub Bot logged work on HIVE-26167:
-----------------------------------------

                Author: ASF GitHub Bot
            Created on: 22/Apr/22 11:46
            Start Date: 22/Apr/22 11:46
    Worklog Time Spent: 10m 
      Work Description: marton-bod commented on PR #3234:
URL: https://github.com/apache/hive/pull/3234#issuecomment-1106434642

   The patch looks good, but one question: when we send in a statement via SQL 
in the normaly way, where does the query state get initialized?




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

    Worklog Id:     (was: 760755)
    Time Spent: 20m  (was: 10m)

> QueryStateMap in SessionState is maintained correctly
> -----------------------------------------------------
>
>                 Key: HIVE-26167
>                 URL: https://issues.apache.org/jira/browse/HIVE-26167
>             Project: Hive
>          Issue Type: Improvement
>            Reporter: László Pintér
>            Assignee: László Pintér
>            Priority: Major
>              Labels: pull-request-available
>          Time Spent: 20m
>  Remaining Estimate: 0h
>
> When the Driver is the QueryStateMap is also initialized with the query ID 
> and the current queryState object. This record is kept in the map until the 
> execution of the query is completed. 
> There are many unit tests that initialise the driver object once during the 
> setup phase, and use the same object to execute all the different queries. As 
> a consequence, after the first execution, the QueryStateMap will be cleaned 
> and all subsequent queries will run into null pointer exception while trying 
> to fetch the current querystate from the SessionState. 



--
This message was sent by Atlassian Jira
(v8.20.7#820007)

Reply via email to