You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@flink.apache.org by neha goyal <ne...@gmail.com> on 2023/02/26 18:23:01 UTC

Is there any API method for dynamic loading of the UDF jar

Hello,

In Flink 16, CREATE FUNCTION USING JAR functionality has been introduced
where we can specify the jar resources and the jar can be located in a
remote file system such as hdfs/s3. I don't see an alternative method for
the same functionality using the TableEnvironment methods call, for
example, createTemporarySystemFunction doesn't take any URI.

Will these methods be provided in the future?
Is there any difference in performance if we use TableEnvironment method
calls vs TableEnvironment.executeSql for the same feature? which one is
recommended?

Thanks and regards

Re: Is there any API method for dynamic loading of the UDF jar

Posted by yuxia <lu...@alumni.sjtu.edu.cn>.
Flink 1.17 which is to be released recently will support it[1] in table environment[2]. 

No any difference in performance between TableEnvironment method calls and TableEnvironment.executeSql, the api is different but the runtime is same under the hood. You can choose any one of them according to your need / perference. 

[1] https://issues.apache.org/jira/browse/FLINK-27660 
[2]: https://github.com/apache/flink/blob/56b124bcfd661a295ab8772d265c12de25f690ab/flink-table/flink-table-api-java/src/main/java/org/apache/flink/table/api/TableEnvironment.java#L541 

Best regards, 
Yuxia 


发件人: "neha goyal" <ne...@gmail.com> 
收件人: "User" <us...@flink.apache.org> 
发送时间: 星期一, 2023年 2 月 27日 上午 2:23:01 
主题: Is there any API method for dynamic loading of the UDF jar 

Hello, 
In Flink 16, CREATE FUNCTION USING JAR functionality has been introduced where we can specify the jar resources and the jar can be located in a remote file system such as hdfs/s3. I don't see an alternative method for the same functionality using the TableEnvironment methods call, for example, createTemporarySystemFunction doesn't take any URI. 

Will these methods be provided in the future? 
Is there any difference in performance if we use TableEnvironment method calls vs TableEnvironment.executeSql for the same feature? which one is recommended? 

Thanks and regards