[ 
https://issues.apache.org/jira/browse/KYLIN-6124?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=18147041#comment-18147041
 ] 

Longfei Jiang commented on KYLIN-6124:
--------------------------------------

[[email protected]] Using system-level configurations for validation is 
unreasonable:
JDBC data sources only need to be configured at the project level to be used, 
and different projects can be configured with different data sources. Using 
system-level configurations cannot simultaneously validate multiple data 
sources.
In the new PR: https://github.com/apache/kylin/pull/2360, this was modified to: 
Check the class itself instead: it must implement IPushDownRunner / 
IPushDownConverter.

> Potential JDBC validation gap in query pushdown configuration
> -------------------------------------------------------------
>
>                 Key: KYLIN-6124
>                 URL: https://issues.apache.org/jira/browse/KYLIN-6124
>             Project: Kylin
>          Issue Type: Bug
>    Affects Versions: 5.0.5
>            Reporter: liyang
>            Assignee: liyang
>            Priority: Major
>             Fix For: 5.0.6
>
>
> Project-level query pushdown configuration accepts JDBC URL, driver, and 
> runner class values without applying the validation used for source JDBC 
> connections. A project administrator can cause pushed-down queries to load a 
> chosen JDBC driver and connect to an arbitrary URL.
> Fix direction: apply the existing JDBC URL and driver allowlist to query 
> pushdown connection properties, and restrict project-level pushdown runner 
> classes to supported implementations.



--
This message was sent by Atlassian Jira
(v10.3.26#10030026)

Reply via email to