Scale Logstash config for more than 20,000 messages per minute?

@leandrojmp, sorry.... it was duplicate reply from my site that why I'm not sure what was provide.
I use persistent queue already, but it the same result. the log status is the same as in this topic: Time delay between logstash and elasticsearch

Elastic host running on SSD of course.

here is my logstash pipeline.yml

- pipeline.id: main
  path.config: "/etc/logstash/conf.d/*.conf"
  pipeline.workers: 20

I will try to follow indexing speed as your recommendation.

Thanks a lot
Trung