TPS <20 even with MSPT <50
Just went through and successfully improved server performance by redoing all my configs. Unfortunatly this seems to have caused or uncovered a different issue, I'm getting <20 TPS even when MSPT is under 50!
https://spark.lucko.me/UwE1XofNSh
Any one know what could be causing this? My google searches were inconclusive.
(before anyone mentions it we are planning on updating to 1.21 soon)
spark
spark is a performance profiler for Minecraft clients, servers, and proxies.
30 Replies
Thanks for asking your question!
Make sure to provide as much helpful information as possible such as logs/what you tried and what your exact issue is
Make sure to mark solved when issue is solved!!!
/close
!close
!solved
!answered
Requested by thedeafcreeper#0
jeezus
yeah...
why is there 1.4k item frames
idk
there's not that many grouped up though at all
its just a lot of people labling their chests spread through the world
(I also implemented some limits on item frames though the other day as well)
also like 100 of the 450 wolves are owned by 1 person 💀
ðŸ˜
tell him to stop god damn
hooooly
yeah I had a lot of lag with a bunch of item frames being loaded and ended up switching to using storage mods instead
how did you limit item frames?
also yeah that is ....a lot of entities wow (although I don't know what is normal for that many players)
and armour stands
and armour stands
Modrinth
Insights - Minecraft Plugin
Extremely fast region scans | Find large redstone machines | Limits per chunk / region | Custom region addon api
armor stands are not ticked so idc that much about them
:nooo:
well update for anyone who looks at this, with super low MSPT (low 20s) TPS seems to be pretty constitently
19.96-19.97and another profiler for good measure, initially low TPS is from the initial flood of players after a restart
https://spark.lucko.me/ZfWhDhjNa8
spark
spark is a performance profiler for Minecraft clients, servers, and proxies.
Admincraft Meta typing randomly 👀
A; citizens ew
b; can you run a spark for ticks over 100
@TheDeafCreeper
https://spark.lucko.me/aGyhJwVXBD
You might have figured it out, the current version of one of our custom plugins (its being rewritten) saves stuff to massive files once a second and I guess the drives are not as fast as the old server so are causing micro lag. (That's what I'm assuming anyways)
I'll have to mess around with that tonight.
spark
spark is a performance profiler for Minecraft clients, servers, and proxies.
once per second? what are you doing?
you have 3 spots of mongodb on the main thread and 2 spots of sync file writes
fix that and itll be fine
personally id toss illegalstack
in ajleaderboard, go to the config and turn off the luckperms contexts
id change your code in your plugins to not use protocollib
use packetevents instead
^^^ to all the above
I dont remember exactly why but I think it was just a "temporary solution" to saving on
onDisable not being super consistent. (It was coded 2 years ago)
I'll go through and do everything Java Enthusiast said though, thanks!
!solvedClosed post!
Your post has been marked as solved!
Requested by thedeafcreeper#0
Why not once per minute?
idk
once per second is crazy
It was 2 years ago 💀
Ok it wasnt what we thought it was here, turns out the issue was that I had turned
tps-catchup off in purpur.yml which was causing it for some reason.
If anyone has a similar issue in the futurewha? I always have that off and have never had that issue ever
Tps being low while mspt is fine usually indicated that the issue isnt with the mc server but with the machine
Etc
It happened on 2 different machines both of which should be plenty strong for an empty minecraft server with no plugins and was fixed by enabling tps-catchup on both so idk
(Intel Core i9-13900 and AMD Ryzen 7 3800X)
why should you turn that off?
Kills performance