In the 3.3.7 separated storage and computation version, Flink connector accidentally split Flink into multiple jobs by writing SR

[lukeyan2023]

• Flink version 1.19
• Starrocks version 3.3.7
• flink-connector-starrocks version 1.2.10_flink-1.19

I wrote a Flink program to read data from Kafka and write it to two different tables in StarRocks. After submitting and running it, I found that the Flink program was accidentally split into multiple jobs, and there were duplicate jobs

The code for the insert into section is as follows:

Normally, after submission, there should only be one job, but unexpectedly it was found to be split into multiple jobs：