search

PaperMC / Folia

Fork of Paper which adds regionised multithreading to the dedicated server.
GNU General Public License v3.0
3.54k stars 446 forks source link

Folia block generation error #61

Open xiaoyueyoqwq opened 1 year ago

xiaoyueyoqwq commented 1 year ago

Expected behavior

It should have safely generated normal blocks following my server's packets, because blocks elsewhere on my server were normal

Observed/Actual behavior

But....Okay, look this then you can know the problems image

Steps/models to reproduce

I think this is a completely random error, because most of my server is normal, including the error map I showed has been generated on the Paper server and copied to Folia, but Folia is still in the wrong state

But I observed a problem, at first Folia did load the maps normally, but as the loading maps got bigger and started to freeze from a certain edge, the maps started to get strange and made the normal loading maps wrong, which happened in the ocean, including the rivers along the oceans

Maybe it's an issue with my packets, but I can't 100% guarantee that this is the case, because the rest of the maps on my server are complete and error-free, so I chose to give you feedback

Plugin and Datapack List

plugins: image

datapackets: image

Folia version

[20:20:50 INFO]: Checking version, please wait... [20:20:50 INFO]: This server is running Folia version git-Folia-"b97116b" (MC: 1.19.4) (Implementing API version 1.19.4-R0.1-SNAPSHOT) (Git: b97116b) You are running the latest version

Other

This is a little bit wrong report, in that time server only output these things...

https://paste.ubuntu.com/p/XBJ8P9Ytwh/

titandino commented 1 year ago

I have had a few instances of terrain generation bugs as well. Nothing that looks like this, though. Mine consisted of just chunk borders similar to when you change the seed and try to generate new terrain.

No real reliable reproduction steps I can provide, unfortunately. I can get it to happen relatively reliably when just driving around in a boat. This was happening on an Intel Xeon Phi 7210 if that helps. It has an extremely high core count but the clock speed is pretty low so that could be why it is happening more often on this machine. I notice it doesn't happen nearly as often on my Intel i9-10600k.

xiaoyueyoqwq commented 1 year ago

I have had a few instances of terrain generation bugs as well. Nothing that looks like this, though. Mine consisted of just chunk borders similar to when you change the seed and try to generate new terrain.

No real reliable reproduction steps I can provide, unfortunately. I can get it to happen relatively reliably when just driving around in a boat. This was happening on an Intel Xeon Phi 7210 if that helps. It has an extremely high core count but the clock speed is pretty low so that could be why it is happening more often on this machine. I notice it doesn't happen nearly as often on my Intel i9-10600k.

maybe you are right? My server was running on intel Xeon E5 server, maybe this is problem...

Spottedleaf commented 1 year ago

Are you sure your datapacks don't cause this?

titandino commented 1 year ago

I am not quite sure what causes it. I can't seem to reproduce it when setting the seed manually in the server.properties.