We have a configuration that is set to
druid.indexer.runner.javaOpts=-server -Xmx2g -Duser.timezone=UTC -Dfile.encoding=UTF-8 -Djava.util.logging.manager=org.apache.logging.log4j.jul.LogManager
It’s description makes me think that each of the indexing task will start with 2g of heap but what I am observing is not the same. I have 30 worker threads which should cause 60g of RAM to be used but the RAM usage is not less than 10g. Is my understanding of this option incorrect?