I am running Flink application with 3 jobs running in it. I faced an issue when multiple jobs keeps restarting for some reason, the jobmanager process got killed and started again (I faced this issue when running in Flink 1.8.0 however Im not able to reproduce this in Flink 1.14.3). It starts with no issues but since the PID list is not cleared, the logs are logged into to another file with incremental number (0 to 1. Ex: flink-test-standalonesession-0-localhost.log to flink-test-standalonesession-1-localhost.log). When comes to purging the logs, the logs from the previous instance did not get cleared. Which end up in filling the disk.
My question would be: can Flink be configured to run only 1 instance even in case of failure or gets killed it should not start another process?