What version are you seeing the problem on?

1.18.2

Describe the issue

Run Spark profiler in server with people
See results

latest.log

https://spark.lucko.me/p4J2fXnR6v?hl=16483,20079,26303,32576,34625,36688,38414,40907,42761,44224,45189,46209,46683,47565,48464,49110,49859,50644,51381,52756,53219,53582,54597,54998,55475,55766,56007,56273,56638,56989,57420,57662,57742,57864,57971,76435,6047

According to the profiling you gave Pollen events are taking a tiny fraction of a tick, even after considering all events. It takes up a reasonable amount of time per tick only if listeners are registered to events.

According to the profiling you gave Pollen events are taking a tiny fraction of a tick, even after considering all events. It takes up a reasonable amount of time per tick only if listeners are registered to events.

??? It took 3944 ms which is 78 ticks.

Spark reports the total time is 438027ms, which is 21901 ticks total. That means 3944ms is literally 0.9% of the tick time. It takes more time to tick animal movement (5033ms) than invoke events.

I see what you mean now. Thanks for clarifying

np