MeshRef is leaking memory causing lag spikes

[nillawafers4u]

Game Version

v1.20.3

Platform

Windows

Modded

Modded

SP/MP

Multiplayer

Description

for some reason myself and one other player on my server have random ping spikes, anywhere from 100ms to 1000ms. another play on the server has no issues whatsoever when this is going on. The server runs on a dedicated machine and monitoring the resources seems to be doing just fine even in the moments this happens. Furthermore I have run a continual ping from my machine to the server computer using the same local IP i connect with and the response time never goes beyond 50 at the most. I have noticed some oddities in my logs whenever this happens such as:

11.2.2025 10:17:36 [Debug] MeshRef with vao id 46326 with 252 indices is leaking memory, missing call to Dispose. Allocated at    at System.Environment.get_StackTrace()
   at Vintagestory.Client.NoObf.ClientPlatformWindows.UploadMesh(MeshData data) in C:\Users\Tyron\Documents\vintagestory\game\VintagestoryLib\Client\ClientPlatform\GLMeshBuffer.cs:line 765
   at Vintagestory.Client.RenderAPIBase.UploadMultiTextureMesh(MeshData data) in C:\Users\Tyron\Documents\vintagestory\game\VintagestoryLib\Client\API\RenderAPIBase.cs:line 86
   at Vintagestory.GameContent.EntityShapeRenderer.<>c__DisplayClass46_1.<TesselateShape>b__1() in C:\Users\Tyron\Documents\vintagestory\game\VSEssentials\EntityRenderer\EntityShapeRenderer.cs:line 283
   at Vintagestory.Client.NoObf.ClientMain.ExecuteMainThreadTasks(Single deltaTime) in C:\Users\Tyron\Documents\vintagestory\game\VintagestoryLib\Client\ClientMain.cs:line 775
   at Vintagestory.Client.GuiScreenRunningGame.RenderToPrimary(Single dt) in C:\Users\Tyron\Documents\vintagestory\game\VintagestoryLib\Client\MainMenu\Screens\GuiScreenRunningGame.cs:line 170
   at Vintagestory.Client.ScreenManager.Render(Single dt) in C:\Users\Tyron\Documents\vintagestory\game\VintagestoryLib\Client\ScreenManager.cs:line 712
   at Vintagestory.Client.ScreenManager.OnNewFrame(Single dt) in C:\Users\Tyron\Documents\vintagestory\game\VintagestoryLib\Client\ScreenManager.cs:line 656
   at Vintagestory.Client.NoObf.ClientPlatformWindows.window_RenderFrame(FrameEventArgs e) in C:\Users\Tyron\Documents\vintagestory\game\VintagestoryLib\Client\ClientPlatform\GameWindow.cs:line 120
   at OpenTK.Windowing.Desktop.GameWindow.Run()
   at Vintagestory.Client.ClientProgram.Start(ClientProgramArgs args, String[] rawArgs) in C:\Users\Tyron\Documents\vintagestory\game\VintagestoryLib\Client\ClientProgram.cs:line 334
   at Vintagestory.Client.ClientProgram.<>c__DisplayClass10_0.<.ctor>b__1() in C:\Users\Tyron\Documents\vintagestory\game\VintagestoryLib\Client\ClientProgram.cs:line 130
   at Vintagestory.ClientNative.CrashReporter.Start(ThreadStart start) in C:\Users\Tyron\Documents\vintagestory\game\VintagestoryLib\Client\ClientPlatform\ClientNative\CrashReporter.cs:line 93
   at Vintagestory.Client.ClientProgram..ctor(String[] rawArgs) in C:\Users\Tyron\Documents\vintagestory\game\VintagestoryLib\Client\ClientProgram.cs:line 130
   at Vintagestory.Client.ClientProgram.Main(String[] rawArgs) in C:\Users\Tyron\Documents\vintagestory\game\VintagestoryLib\Client\ClientProgram.cs:line 48
   at Vintagestory.ClientWindows.Main(String[] args) in C:\Users\Tyron\Documents\vintagestory\game\VintagestoryClient\Windows\ClientWindows.cs:line 9.

The strangest part to me is that as stated before it only affects some players connecting. Where they connect from seems to make no difference as even I am affected when connecting over my local network.

How to reproduce

the only thing that seems to reproduce this ping spikes is moving around in the world. I think its anytime a new block is rendered, as staying completely still seems to stop it from occurring.

Screenshots

No response

Logs

client-debug.log

client-main.log

[nillawafers4u]

forgot to mention as well I run the game on a windows 11 PC with 32gb of ram. My hardware doesn't really seem to be making out when these spike happen either ( I had my resource manager open while testing to see)

[Craluminum2413]

Caused by some mod. I can't reproduce it in vanilla

[nillawafers4u]

I believe this might be something in vanilla since it still persists without any mods.

I will attach the logs from my game after doing a complete reinstall. The same errors and ping spikes persist. Also I'd also like to note that this issue only started occurring after updating from 1.20.1 to 1.20.3. downgrading does not seem to fix the issue though. Maybe I'm still missing something. Just thought I'd provide more information.

client-debug.log
client-main.log

[Craluminum2413]

Enable these settings, load the game then send logs again

[nillawafers4u]

As requested. I went ahead and wiped the server I'm hosting too before getting these new logs, so fresh world and all too.

client-debug.log
client-main.log

[Craluminum2413]

@nillawafers4u Do you get these errors in normal world?

[Craluminum2413]

Ok I reproduced it

[thelyric27]

I was experiencing this issue too, causing spikes of lag and also freezing chunk loading in my world, it was really frustrating. MeshRef with vao id 40899 with 1932 indices. But even more frustratingly, I couldn't replicate the issue after turning on debug logging. In the world I was playing the chunks still wouldn't load but the meshref error never showed up in the log. I tried two other worlds and in one, traveled 8000 blocks from spawn thinking it was some particular block or item causing it, but suddenly chunks would load smoothly and without delay, I could sprint in one direction for in-game hours.