In general what I saw is that to reduce the log volume is that people only send error or warning log lines instead of also info, so the critical information still reaches the elasticsearch cluster. So I would recommend you to look a way to reduce the amount of logs based on some criterias instead of just a percentage.
I'm not aware that any of the other products can do that but you must check nxlog and the others in detail to see if they support it.
Apache, Apache Lucene, Apache Hadoop, Hadoop, HDFS and the yellow elephant
logo are trademarks of the
Apache Software Foundation
in the United States and/or other countries.