[
https://issues.apache.org/jira/browse/KYLIN-5983?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17927956#comment-17927956
]
Guoliang Sun commented on KYLIN-5983:
-------------------------------------
Fix Design:
* Disable Gluten functionality for asynchronous queries (which have a separate
queue).
* Remove Gluten-related configurations during asynchronous queries.
* Turn off the internal table switch and remove the internal table catalog.
* Since the metadata database is used to initialize the config in asynchronous
queries, modify the query matching logic to remove the internal table matching
logic.
> Asynchronous query exception
> ----------------------------
>
> Key: KYLIN-5983
> URL: https://issues.apache.org/jira/browse/KYLIN-5983
> Project: Kylin
> Issue Type: Bug
> Affects Versions: 5.0.0
> Reporter: Guoliang Sun
> Assignee: Guoliang Sun
> Priority: Major
> Attachments: image-2025-02-18-15-53-28-895.png
>
>
> RC: After Kylin integrates Gluten and internal tables, a task type judgment
> logic was added in the Spark application. This ensures that the loading tasks
> for internal tables will enable Gluten, while other build tasks will not
> enable Gluten. The code is as follows:
> !image-2025-02-18-15-53-28-895.png|width=878,height=111!
> Asynchronous queries also reuse this logic, but since the *P_CLASS_NAME* is
> not specified for asynchronous queries, it leads to a null pointer issue.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
