Indexing tasks fail intermittently without any error

My Kafka indexing tasks fail intermittently i.e. there is no clear pattern. It fails once and the task gets picked up in another node or even the same node and it succeeds. After a few successful runs, it fails again. From the indexing log, I don’t see any error. It actually fails even before it begins to read records and begin processing. I have attached the log file here but the last few lines indicate that JVM gets shutdown after acquiring the lock.

2020-04-15T22:44:21,885 INFO [main] org.apache.druid.indexing.worker.executor.ExecutorLifecycle - Attempting to lock file[var/druid/task/index_kafka_hour_level_druid_ingestion_test_b5e018d2e1bdca4_ailnfdbi/lock].

2020-04-15T22:44:21,888 INFO [main] org.apache.druid.indexing.worker.executor.ExecutorLifecycle - Acquired lock file[var/druid/task/index_kafka_hour_level_druid_ingestion_test_b5e018d2e1bdca4_ailnfdbi/lock] in 3ms.

2020-04-15T22:44:21,889 INFO [parent-monitor-0] org.apache.druid.indexing.worker.executor.ExecutorLifecycle - Triggering JVM shutdown.

``

index_task.log (53.7 KB)