Currently we print all spark launcher's log to fe.log. If multiple spark load jobs are submitted at the same time, their logs will be confused and difficult to distinguish.
Therefore, I suggest that the log of each spark load should be separated and written into a single log file. These logs will be saved in the fe/log/spark-launcher-log/ directory until the old label is cleared.
Currently we print all spark launcher's log to fe.log. If multiple spark load jobs are submitted at the same time, their logs will be confused and difficult to distinguish.
Therefore, I suggest that the log of each spark load should be separated and written into a single log file. These logs will be saved in the
fe/log/spark-launcher-log/directory until the old label is cleared.