Hi, I am trying to load data from postgresql tables through Table APIs. I am using BATCH as execution mode. I am trying to load data from postgresql table, aggregate it and store it back to sink. I have kept 8 as parallelism, 8 task slots and 1 task manager. When I try to load data I could see that 7/8 subtasks get finished almost instantaneously and 1/8 subtask runs for extended period of time to load data from the source. I am not understanding why is this happening. Could someone please help? Is there any configuration which I am missing? P.S I am using Flink 1.15