[
https://issues.apache.org/jira/browse/HUDI-6832?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Ma Jian updated HUDI-6832:
--------------------------
Description:
The current multi table service supports executing table service for multiple
tables by specifying them through {{{}hoodie.tableservice.tablesToServe{}}}.
When an incorrect path is provided for a table or a table is deleted but the
asynchronous table service is not yet aware of it, the current implementation
causes the entire Spark job to fail, impacting the execution of other tables.
It is expected that this should not impact the table service of other tables
with correctly configured paths.
!image-2023-09-08-14-49-44-564.png|width=840,height=111!
In order to address this issue, I will introduce a check to skip such cases and
log the details of tables that do not conform to Hudi's requirements using
Log.error.
was:
The current multi table service supports executing table service for multiple
tables by specifying them through {{{}hoodie.tableservice.tablesToServe{}}}.
When an incorrect path is provided for a table or a table is deleted but the
asynchronous table service is not yet aware of it, the current implementation
causes the entire Spark job to fail, impacting the execution of other tables.
It is expected that this should not impact the table service of other tables
with correctly configured paths.
!image-2023-09-08-14-49-44-564.png!
In order to address this issue, I will introduce a check to skip such cases and
log the details of tables that do not conform to Hudi's requirements using
Log.error.
> When one of table paths is incorrect, ensure that other table services are
> not affected.
> ----------------------------------------------------------------------------------------
>
> Key: HUDI-6832
> URL: https://issues.apache.org/jira/browse/HUDI-6832
> Project: Apache Hudi
> Issue Type: Bug
> Reporter: Ma Jian
> Priority: Minor
> Attachments: image-2023-09-08-14-49-44-564.png
>
>
> The current multi table service supports executing table service for multiple
> tables by specifying them through {{{}hoodie.tableservice.tablesToServe{}}}.
> When an incorrect path is provided for a table or a table is deleted but the
> asynchronous table service is not yet aware of it, the current implementation
> causes the entire Spark job to fail, impacting the execution of other tables.
> It is expected that this should not impact the table service of other tables
> with correctly configured paths.
> !image-2023-09-08-14-49-44-564.png|width=840,height=111!
> In order to address this issue, I will introduce a check to skip such cases
> and log the details of tables that do not conform to Hudi's requirements
> using Log.error.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
