[Workflow] Shell task (beeline -f) used wrong value in actual execution.

[pppeace]

In past days, this issue happened twice.

Here are some logs fragment

[LOG-PATH]: /home/hadoop/module/dolphinscheduler/worker-server/logs/20240607/13421671136576_15-3879-13096.log, [HOST]:  Host{address='10.122.130.109:1234', ip='10.122.130.109', port=1234}
[INFO] 2024-06-07 16:12:38.889 +0800 - Begin to pulling task
[INFO] 2024-06-07 16:12:38.890 +0800 - Begin to initialize task
[INFO] 2024-06-07 16:12:38.890 +0800 - Set task startTime: Fri Jun 07 16:12:38 CST 2024
[INFO] 2024-06-07 16:12:38.890 +0800 - Set task envFile: /home/hadoop/module/dolphinscheduler/worker-server/conf/dolphinscheduler_env.sh
[INFO] 2024-06-07 16:12:38.890 +0800 - Set task appId: 3879_13096
[INFO] 2024-06-07 16:12:38.890 +0800 - End initialize task
[INFO] 2024-06-07 16:12:38.890 +0800 - Set task status to TaskExecutionStatus{code=1, desc='running'}
[INFO] 2024-06-07 16:12:38.891 +0800 - TenantCode:hadoop check success
[INFO] 2024-06-07 16:12:38.891 +0800 - ProcessExecDir:/home/hadoop/module/dolphinscheduler/data/exec/process/hadoop/12835742987584/13421671136576_15/3879/13096 check success
[INFO] 2024-06-07 16:12:38.891 +0800 - get resource file from path:/dolphinscheduler2/hadoop/resources/xxx.sql
[INFO] 2024-06-07 16:12:38.916 +0800 - Resources:{xxx.sql=hadoop} check success
[INFO] 2024-06-07 16:12:38.917 +0800 - Task plugin: SHELL create success
[INFO] 2024-06-07 16:12:38.917 +0800 - shell task params {"localParams":[],"rawScript":"#!/bin/bash\r\n\r\n/home/hadoop/spark-3.1.1-bin-hadoop3.2/bin/beeline -u jdbc:hive2://node2:11240/default -n hive -p hive -f xxx.sql --hivevar ads ${ads} --hivevar dim ${dim} --hivevar partition_version ${partition_version} --hivevar dws ${dws} --hivevar dwd ${dwd}","resourceList":[{"id":40,"resourceName":"ads_sub_pn_sub_region_ym_ur.sql","res":"xxx.sql"}]}
[INFO] 2024-06-07 16:12:38.917 +0800 - Success initialized task plugin instance success
[INFO] 2024-06-07 16:12:38.917 +0800 - Success set taskVarPool: null
[INFO] 2024-06-07 16:12:38.917 +0800 - raw script : #!/bin/bash

/home/hadoop/spark-3.1.1-bin-hadoop3.2/bin/beeline -u jdbc:hive2://node2:11240/default -n hive -p hive -f ads_sub_pn_sub_region_ym_ur.sql --hivevar ads ads_test --hivevar dim dim_tmp --hivevar partition_version default_latest --hivevar dws dws_test --hivevar dwd dwd_tmp
#################################################
	Error: org.apache.hive.service.cli.HiveSQLException: Error running query: org.apache.spark.sql.AnalysisException: `ads_partition_test`.`ads_sub_pn_sub_region_ym_ur` requires that the data to be inserted have the same number of columns as the target table: target table has 43 column(s) but the inserted data has 42 column(s), including 1 partition column(s) having constant value(s).

Actually, the table structure is not same at the different database ads_test and ads_partition_test. And in the previous log, the hivevar value is correct(its ads_test), but in the execution, it turns to wrong value(ads_partition_test).

Any idea about this issue?