Throughput issue on ubuntu 18

[ResearchIntern98]

Running ipfixcol on ubuntu 16, produces 0.1 million kafka record per second, while running the same produces 17-20k kafka record per second.

[thorgrin]

You will have to be much more specific if you expect the issue to be answered. Ubuntu 16 compared to what? What version of the kafka library ale you using? Which server? How does your configuration file look like? What are the steps to reproduce the measurement? Can you prepare a Vagrant file(s) to demonstrate the difference?

[ResearchIntern98]

Kafka version 2.11-2.3.0. Running ipfixcol on ubuntu 16, produces 0.1 million kafka record per second, while running the same on ubuntu18 produce 17-20k kafka record per second. Machines are Dell servers. On kafka cluster, we run kafka console consumer for whole cluster and measure the number of records consumed within specified time for that topic.