Epoch Server Performance

[C4-timah]

I could use some recommendations on gaining some performance back on my server. Currently the CPU with no one on the server runs between at a baseline of 26% while jumping to the the 63% range frequently every couple of minutes or so. Then when players hop on, the baseline percentage jumps to 42% and up. The following addons are currently running:

@a3_epoch_WAI (default settings), @epochz (default settings), @ryanzombies, @HS_Traders.

I'm sure having roaming Z's and Roaming AI with missions is most definitely the cause, but what would be a good mix of triggered spawns vs. roaming? Or should I simply remove triggered all together? I'm looking for a good balance if its possible. I am new at the Arma Dedicated Server game so please be patient with me. Your suggestions and tips are greatly appreciated!

Edited October 11, 2018 by C4-timah
default addon settings

[C4-timah]

Apparently editing epochConfig.hpp by commenting out some of the default spawning events (Carnival, Money Drop, Plant Spawner, Earthquake, etc.)   and reducing the epoch traders ( NPCSlotsLimit = 7 ) from the default amount has dropped the idle cpu % down by almost 20%!

You can see where the CPU usage drops considerably when I make the change and restart. There is an increase when 3 players join and drops back to the 20% range when the leave.

Edited October 12, 2018 by C4-timah

[Helion4]

Ryans Zombies is a massive performance hit, according to every server host that has ever used it. They say it is badly optimised. Maps can also affect performance sometimes, hosts say the Chern. Redux is a performance hog, for example.

Personally I would not be bothered by those numbers, and only really be worried when players come to you with reports of server lag -

(which is VERY different from client lag, and often includes things like AI rubber banding over 30 meters, or building objects not saving because the Database is struggling)

- or stuttering or worse, like not being able to connect to the server at all.
 

for more info, you could repeat the test with only Epoch active on Altis, this way you get baseline and you modded server results.