1.16 Server Optimisation Help

Hey! I have just done “some sort” of stress test on my 1.16 server. Yes it was just 80 ish bots jumping around, running around in clumps etc but it gives you a good idea of how things COULD go. Im just wondering if theres any more optimisation that could be suggested, or plugin changes for better plugins.

Timings:
https://timings.aikar.co/?id=156031aa5e954159b2a8befe45a62884#timings

I will be adding some sort of bungee queue to try reduce the effect of mass joinings on the server. The server has 350% cores wise & 13GB RAM. If theres nothing else i can really do then thats fine, but it doesn’t hurt to ask. :slight_smile:

Thanks, Shep

I’m wondering how you stress tested it? Is there a program you used to get bots to connect?

There are three things which sticks out:

  • tracker stage 2 - not sure what this is but on my server it only uses 5% of the tick as opposed to your 101.94%. If you have the plugin “Simplebackup” or another backup plugin, this might be the cause.

  • Connection Handler - Seems like a combination of many plugins so it might be difficult to fix.

  • nms.PacketPlayInFlying$PacketPlayInPosition - Some listeners for different events it would seem, part of a variety of different plugins.

Also, Worldguard is using 22% of tick which is very high for a single plugin, have had the same problem with WorldGuard on my server. Not sure how I solved it or if there is a solution. Straffreport takes 10% of the TPS which is way more than a plugin with that functionaly should use. That indicates a poorly optimised plugin, maybe with a memory leak.