Hi all, the compaction task and automatic compaction are being used for optimizing generated segments in background. They are especially useful when it comes to the realtime ingestion, because the segments generated by realtime tasks are not really optimized.
However, with the current timeChunk locking and segment versioning system, it's not possible to run compactionTasks while ingesting data into the same timeChunk (https://github.com/apache/incubator-druid/issues/6136). https://github.com/apache/incubator-druid/issues/6319 is proposing a new locking and versioning system to fix this issue. I'd like to draw attention from more people since it would be a huge change. Please take a look and let me know what you think. Best regards, Jihoon