A test to compare performance between using ConcurrentLinkedQueue & LMAX disruptor.
The test are to send 50M messages to be consumed.
In Vanilla test, it uses ConcurrentLinkedQueue, starts 20 parallel producers & consumers, while LMAX it uses one producer and multiple subscribers.
I see substantial improvement using LMAX (1.8 vs 20.56 secs).
Time to switch to LMAX then :)