Why would you have 120 workers configured per pipeline? That sounds excessive and I would recommend reducing that significantly so the total number of workers roughly matches the number of CPU cores on the host where it runs. Larger number does not necessarily equal better performance.
If you want to improve performance instead look at increasing the batch size a bit but do not go crazy.
There is no hard limit with respect to how many connections a node can handle as it depends on the load and how many indices and shards each bulk requests targets. You will need to test.
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.