[1.10.2] control panel cause tps drops, chunk issues
Opened this issue ยท 13 comments
I wasn't able to replicate this so far.
- What changes did you make to the modpack since the crashes in #11?
- Does this happen on a non-sponge server/in singleplayer?
- Does the chunk unload if a) there is not control panel or b) if the indicator is off?
- How heavy is heavy load? Something I'd absolutely notice while testing or something I'd have to look at the memory/CPU usage of the server for?
- only IW updated
- tested only on the sponge server
- If the indicator is off no problem - chunk is unloaded
- tps drops to <5
Btw i made some others tests and update sponge server too. It looks like problem solved. Please dont close issue, I will make some other tests, then report.
With latest spongeforge build 2555 (chunk unload optimalizations) unloading chunks with panel
https://www.youtube.com/watch?v=n_E7DQwfF2Y&index=1&ab_channel=HonzaVencl%C5%AF
Indicator must be light on
cause crash server with:
Description: Exception ticking world entities java.util.ConcurrentModificationException at java.util.ArrayList.removeIf(ArrayList.java:1411) at net.minecraft.world.World.removeTileEntitiesForRemovedChunks(World.java:5669) at net.minecraft.world.World.func_72939_s(World.java:5610) at net.minecraft.world.WorldServer.func_72939_s(WorldServer.java:2185) at net.minecraft.server.MinecraftServer.func_71190_q(MinecraftServer.java:709) at net.minecraft.server.dedicated.DedicatedServer.func_71190_q(DedicatedServer.java:387) at net.minecraft.server.MinecraftServer.func_71217_p(MinecraftServer.java:613) at net.minecraft.server.MinecraftServer.run(MinecraftServer.java:471) at java.lang.Thread.run(Thread.java:745) A detailed walkthrough of the error, its code path and all known details is as follows: --------------------------------------------------------------------------------------- -- Head -- Thread: Server thread Stacktrace: at java.util.ArrayList.removeIf(ArrayList.java:1411) at net.minecraft.world.World.removeTileEntitiesForRemovedChunks(World.java:5669) at net.minecraft.world.World.func_72939_s(World.java:5610) at net.minecraft.world.WorldServer.func_72939_s(WorldServer.java:2185) -- Affected level -- Details: Level name: skyworld All players: 1 total; [EntityPlayerMP['HunterzCZ'/74, l='skyworld', x=-260.36, y=148.00, z=-246.71]] Chunk stats: ServerChunkCache: 856 Drop: 0 Level seed: -7299770919580520601 Level generator: ID 00 - default, ver 1. Features enabled: true Level generator options: Level spawn location: World: (8,64,8), Chunk: (at 8,4,8 in 0,0; contains blocks 0,0,0 to 15,255,15), Region: (0,0; contains chunks 0,0 to 31,31, blocks 0,0,0 to 511,255,511) Level time: 38898578 game time, 25645799 day time Level dimension: 0 Level storage version: 0x04ABD - Anvil Level weather: Rain time: 37009 (now: false), thunder time: 29906 (now: false) Level game mode: Game mode: survival (ID 0). Hardcore: false. Cheats: true -- Sponge CauseTracker -- Details: Cause Stack: [Empty stack] Stacktrace: at net.minecraft.server.MinecraftServer.handler$onCrashReport$zhp000(MinecraftServer.java:3956) at net.minecraft.server.MinecraftServer.func_71230_b(MinecraftServer.java:829) at net.minecraft.server.dedicated.DedicatedServer.func_71230_b(DedicatedServer.java:361) at net.minecraft.server.MinecraftServer.run(MinecraftServer.java:499) at java.lang.Thread.run(Thread.java:745) -- System Details --
with sponge build 2548 (without chunk unload optimalization) panel (setup as above) cause tps drop, because chunks cannot be unloaded
I think I know what is going on, I just have to get SpongeForge working in my development environment.
I started backporting the fix yesterday, but I noticed that the chunks were still not unloading properly. This may be a bug with IE's redstone wire, I'll have to do some more testing today. I didn't notice any lag caused by the chunk not unloading.
I've backported the fix, the unloading issues are caused by some IE code. It is fixed in 1.11+, but Blu will NOT release any more versions below MC 1.12. I can't get Sponge working in my 1.10 dev env, so I can't tell you whether that bug causes more crashes.
I'll publish new builds later today. If you experience any more issues with sponge (I know about at least 2 more) try without Sponge before reporting the bug. Sponge falls into the category of "(K)Cauldron and similar server software" in the contributing guidelines (I'll update that, I wrote those guidelines for IE in 1.7.10). I would probably have closed this if that change in Sponge hadn't been copied from 1.12 Forge, meaning that I would probably have had to deal with it at some point anyway.