Damien
- aeduna
- aeduna
- Europe/Stockholm
- Yes
- No
MC-130959
WEB-1249
server: Linux arch 3.2.0-4-amd64 #1 SMP Debian 3.2.46-1+deb7u1 x86_64 GNU/Linux
client: Linux temple 3.10-2-amd64 #1 SMP Debian 3.10.7-1 (2013-08-17) x86_64 GNU/Linuxjava: Sun Java Platform Standard Edition Runtime Environment (JRE) 6
on bothclient:
System.getProperty('java.version') == '1.6.0_26'
System.getProperty('java.vendor') == 'Sun Microsystems Inc.'
System.getProperty('sun.arch.data.model') == '32'
server: Linux arch 3.2.0-4-amd64 #1 SMP Debian 3.2.46-1+deb7u1 x86_64 GNU/Linux
client: Linux temple 3.10-2-amd64 #1 SMP Debian 3.10.7-1 (2013-08-17) x86_64 GNU/Linuxjava: Sun Java Platform Standard Edition Runtime Environment (JRE) 6
on bothUpdated
server:
java version "1.7.0_40"
Java(TM) SE Runtime Environment (build 1.7.0_40-b43)
Java HotSpot(TM) 64-Bit Server VM (build 24.0-b56, mixed mode)client:
System.getProperty('java.version') == '1.6.0_26'
System.getProperty('java.vendor') == 'Sun Microsystems Inc.'
System.getProperty('sun.arch.data.model') == '32'
I've taken a 1.12.2 world and tried to start it up under 1.13.pre6.
It starts ok, and runs hot (300%+ CPU) for 2 minutes, eventually outputting
[19:30:01] [Server thread/INFO]: Preparing spawn area: 100%
[19:30:01] [Server thread/INFO]: Time elapsed: 125724 ms
[19:30:01] [Server thread/INFO]: Done (126.673s)! For help, type "help"
But when I login, it spikes back up to 350% CPU, and takes a while to show terrain. I didn't move, just sat and waited for the CPU to drop back off, but eventually it crashed.
I ran the same world under pre5, and I could login and move around without a crash (suffered from TPS issues, but not this)
server crash - datafixers issue?server crash - datafixer / loading existing chunks
server crash - datafixer / loading existing chunks / WorldGen-Worker
debian testing
open jdk 16.0.1
01:00.0 VGA compatible controller: NVIDIA Corporation GK104 [GeForce GTX 680] (rev a1)debian testing
open jdk 16.0.1device: 01:00.0 VGA compatible controller: NVIDIA Corporation GK104 [GeForce GTX 680] (rev a1)
driver: nvidia-current-460.73.01
Damien I replied this this comment:
and I know what the problem is in this ticket, I know it has to do with old maps, I'm just telling him that I don't have the issue at all, not even once saw it@FVbico sorry i'm french, i don't understand your answer, are you saying you have no errors and no lags on your map?
I have a similar problem with a crowded underground chicken bunker - if you step just outside, logout, then log back in, you can hear chickens squawking lots and on reentry there are a number of corpses & feathers along the walls. No particular side is favoured.
Still happening in 13w43a - not really feeling the love for being close to a pre-release
---- Minecraft Crash Report ----
// Don't do that.
Time: 22/10/13 12:57 PM
Description: Exception ticking world
java.lang.IllegalArgumentException: n must be positive
at java.util.Random.nextInt(Random.java:300)
at agk.a(SourceFile:122)
at agk.a(SourceFile:27)
at age.a(SourceFile:362)
at agp.a(SourceFile:138)
at apj.a(SourceFile:432)
at mi.a(SourceFile:160)
at aoh.a(SourceFile:781)
at mi.c(SourceFile:95)
at mi.d(SourceFile:108)
at afl.e(SourceFile:311)
at mj.g(SourceFile:234)
at mj.b(SourceFile:137)
at net.minecraft.server.MinecraftServer.u(SourceFile:503)
at lj.u(SourceFile:259)
at net.minecraft.server.MinecraftServer.t(SourceFile:441)
at net.minecraft.server.MinecraftServer.run(SourceFile:358)
at ky.run(SourceFile:617)
A detailed walkthrough of the error, its code path and all known details is as follows:
---------------------------------------------------------------------------------------
– Head –
Stacktrace:
at java.util.Random.nextInt(Random.java:300)
at agk.a(SourceFile:122)
at agk.a(SourceFile:27)
at age.a(SourceFile:362)
at agp.a(SourceFile:138)
at apj.a(SourceFile:432)
at mi.a(SourceFile:160)
at aoh.a(SourceFile:781)
at mi.c(SourceFile:95)
at mi.d(SourceFile:108)
at afl.e(SourceFile:311)
at mj.g(SourceFile:234)
at mj.b(SourceFile:137)
– Affected level –
Details:
Level name: world
All players: 1 total; [mm['aeduna'/499, l='world', x=200.50, y=81.50, z=0.50]]
Chunk stats: ServerChunkCache: 928 Drop: 0
Level seed: -5522882376807769887
Level generator: ID 00 - default, ver 1. Features enabled: true
Level generator options:
Level spawn location: World: (120,64,252), Chunk: (at 8,4,12 in 7,15; contains blocks 112,0,240 to 127,255,255), Region: (0,0; contains chunks 0,0 to 31,31, blocks 0,0,0 to 511,255,511)
Level time: 24823860 game time, 24823860 day time
Level dimension: 0
Level storage version: 0x04ABD - Anvil
Level weather: Rain time: 53100 (now: false), thunder time: 70209 (now: false)
Level game mode: Game mode: creative (ID 1). Hardcore: false. Cheats: false
Stacktrace:
at net.minecraft.server.MinecraftServer.u(SourceFile:503)
at lj.u(SourceFile:259)
at net.minecraft.server.MinecraftServer.t(SourceFile:441)
at net.minecraft.server.MinecraftServer.run(SourceFile:358)
at ky.run(SourceFile:617)
– System Details –
Details:
Minecraft Version: 13w43a
Operating System: Linux (amd64) version 3.2.0-4-amd64
Java Version: 1.7.0_40, Oracle Corporation
Java VM Version: Java HotSpot(TM) 64-Bit Server VM (mixed mode), Oracle Corporation
Memory: 773714672 bytes (737 MB) / 1029701632 bytes (982 MB) up to 1029701632 bytes (982 MB)
JVM Flags: 2 total; -Xmx1024M -Xms1024M
AABB Pool Size: 15208 (851648 bytes; 0 MB) allocated, 0 (0 bytes; 0 MB) used
IntCache: cache: 0, tcache: 0, allocated: 13, tallocated: 95
Profiler Position: N/A (disabled)
Vec3 Pool Size: 2608 (146048 bytes; 0 MB) allocated, 0 (0 bytes; 0 MB) used
Player Count: 1 / 20; [mm['aeduna'/499, l='world', x=200.50, y=81.50, z=0.50]]
Is Modded: Unknown (can't tell)
Type: Dedicated Server (map_server.txt)
hunh.
Further investigation suggests that the disconnect (n must be positive) seems to happen only/to happen mostly on a world where a schematic has been copied in from a 1.6.4 world.
I started with a clean world, and was able to travel around ok. Then I pasted this (large) schematic in, and when I approached it, the server crashed with "n must be positive"
I manage to login again and moved away from the pasted area without a crash. I moved in a big circle and approached the area from another direction, and on getting close enough to the border, it crashed again. I logged in again and teleported into the centre of the pasted area and it crashed straight away.
Ok, I took the 1.6.4 world, run it with the snapshot for 13w43a, and then saved to update it to 1.7ish.
Then I took schematic more or less the same as the one that had been saved from the 1.6.4 server, and copied it into a clean 13w43a server. No crashes from "n must be positive" at all.
So I guess the problem is chunks of 1.6 worlds copied into 1.7?
I can confirm it happens on 1.9 release. I've got a whole village of villagers constantly opening and closing doors they then can't get through. Its... loud.
I'm not getting any crashes, but the daily logs are now 10X the size they were before 1.9 was released. This is impacting how much space each server takes up.
Selected entities keep reporting this message to the console, mostly persistent ones such as villagers and golems, whenever players move around and load chunks.
I can see a huge cascade whenever a player takes a rail past a village for instance.
"I don't have the issue = I don't have the errors at all"
I might be missing something. Why would you bother commenting that? I don't know that its a universal problem, so I don't think everyone will have it?
It looks like a problem with old maps being brought in - mobs that are persistent for some reason (villagers, or zombies holding things, or named, etc) are spamming the logs. Not everyone sees a slow down.
Ditto on multiplayer - zooming through the underdark in spectator mode is amusing.
Multiplayer. I notice the sinking effect, but i was in creative and thought it was a new feature for that mode.
Seems to be resolved....
its, uh, not more informative....
latest.log
I could probably pull out the particular region and attach that if it'd help...
This isn't really a resolution, but I took backups of the offending regions and restored them.
Started with 2G and it crashed again
Started with 4G and managed to login, move about, then quit.
Restarted with 2G and I've managed to login.
.... but 30s later i couldn't interact with things properly, and its using 390% of CPU and then crashed again
[18:51:52] [WorldGen-Scheduler-1/ERROR]: Something went wrong upgrading!
java.lang.OutOfMemoryError: GC overhead limit exceeded
so yah.
Crash report and log file crash-2018-06-06_21.43.41-server.txt
DisableExplicitGC - nope, sorry
Immediately jumps up to 390% cpu as soon as I login and crashes ~ 60s later. (let me know if I should attach logs from that run too)
Not gunna lie, wish there was a seperate utility for converting worlds to 1.13 so that it was clearer where issues were, in the conversion or the generation of new chunks.
This has just started on a world that I've been building on for a few weeks with snapshots - I wandered off in a new direction and it booted me and crashed. Its possible I'd already been that way with one of the older snapshots and so they were already generated chunks. :-|
I can't login now, it crashes instantly. Might try rolling back a version.
EDIT: yep, 18w22c worked ok, ran away from the problem chunks and I can start 1.13.pre1 again now. I guess... I find a different swamp?
Not gunna lie, thought this was on purpose, and I was sort of enjoying it
Dunno, pre5 was fine.
Well.. it suffered from no iron golems and a slow TPS, but I could login and move around, CPU stayed at 50-10% and didn't crash.
I think this is new.
The behaviour in pre7 is a bit different. It still takes 2minutes to go from
[12:18:51] [Server thread/INFO]: Preparing start region for level 0
[12:20:52] [Server thread/INFO]: Preparing spawn area: 0%
but the CPU rans less intensively.
I was able to login but it was very slow in areas where chunks already existed, and eventually crashed. I started the server again, and logged in and it crashed immediately. I got a fresh copy of the world and tried again - I was able to go to the same location (walking instead of flying) and move around without a crash, but after visiting other 1.12.2 generated chunks, it crashed again.
I restarted the server, got another player to login at a different location and it crashed also.
I'll attach various crash logs from today (pre7)
The first crash seems to be an issue with the WorldGen pool? And then maybe its in a weird state and crashes?
Is it much too late to get a utility to just convert the world before starting play? The conversion process is causing endless issues for me as well (eg https://bugs.mojang.com/browse/MC-132768), I'd much rather run something once as a process for updating, even if it takes hours (4+Gig of worl)
pre8 - same behaviour except it took longer to get it to crash. I had to fly around fast in creative in known complicated areas.
Once its crashed once, you cannot login again, because the server times out.
[22:33:56] [Server Watchdog/FATAL]: A single server tick took 60.82 seconds (should be max 0.05)
[22:33:56] [Server Watchdog/FATAL]: Considering it to be crashed, server will forcibly shutdown.
But I fiddled server properties and set the
max-tick-time=600000
to see what might happen.
[22:41:17] [Server thread/ERROR]: Something went wrong upgrading!
java.lang.OutOfMemoryError: Java heap space
at hd.c(SourceFile:50) ~[minecraft_server.pre8.jar:?]
at hd.b(SourceFile:10) ~[minecraft_server.pre8.jar:?]
at gy.f(SourceFile:395) ~[minecraft_server.pre8.jar:?]
at gy.b(SourceFile:28) ~[minecraft_server.pre8.jar:?]
at he.c(SourceFile:271) ~[minecraft_server.pre8.jar:?]
at he.b(SourceFile:17) ~[minecraft_server.pre8.jar:?]
at gy.f(SourceFile:395) ~[minecraft_server.pre8.jar:?]
at gy.b(SourceFile:28) ~[minecraft_server.pre8.jar:?]
at hj.a(SourceFile:160) ~[minecraft_server.pre8.jar:?]
at hj.mergeInto(SourceFile:22) ~[minecraft_server.pre8.jar:?]
at com.mojang.datafixers.types.DynamicOps.set(DynamicOps.java:162) ~[minecraft_server.pre8.jar:?]
at com.mojang.datafixers.types.DynamicOps.lambda$update$11(DynamicOps.java:166) ~[minecraft_server.pre8.jar:?]
at com.mojang.datafixers.types.DynamicOps$$Lambda$1638/1344684320.apply(Unknown Source) ~[?:?]
at java.util.Optional.map(Optional.java:215) ~[?:1.8.0_171]
at com.mojang.datafixers.types.DynamicOps.update(DynamicOps.java:166) ~[minecraft_server.pre8.jar:?]
at com.mojang.datafixers.Dynamic.lambda$update$12(Dynamic.java:151) ~[minecraft_server.pre8.jar:?]
at com.mojang.datafixers.Dynamic$$Lambda$1636/1157730944.apply(Unknown Source) ~[?:?]
at com.mojang.datafixers.Dynamic.map(Dynamic.java:42) ~[minecraft_server.pre8.jar:?]
at com.mojang.datafixers.Dynamic.update(Dynamic.java:151) ~[minecraft_server.pre8.jar:?]
at zu.a(SourceFile:29) ~[minecraft_server.pre8.jar:?]
at zu$$Lambda$547/621337084.apply(Unknown Source) ~[?:?]
at com.mojang.datafixers.DataFix.lambda$null$2(DataFix.java:42) ~[minecraft_server.pre8.jar:?]
at com.mojang.datafixers.DataFix$$Lambda$1468/1907584040.apply(Unknown Source) ~[?:?]
at com.mojang.datafixers.FunctionType$$Lambda$1450/1016980185.apply(Unknown Source) ~[?:?]
at com.mojang.datafixers.FunctionType$Instance.lambda$first$2(FunctionType.java:84) ~[minecraft_server.pre8.jar:?]
at com.mojang.datafixers.FunctionType$Instance$$Lambda$1457/369384969.apply(Unknown Source) ~[?:?]
at com.mojang.datafixers.FunctionType$$Lambda$1450/1016980185.apply(Unknown Source) ~[?:?]
at com.mojang.datafixers.FunctionType$Instance.lambda$null$0(FunctionType.java:76) ~[minecraft_server.pre8.jar:?]
at com.mojang.datafixers.FunctionType$Instance$$Lambda$1456/1517960227.apply(Unknown Source) ~[?:?]
at com.mojang.datafixers.FunctionType$$Lambda$1450/1016980185.apply(Unknown Source) ~[?:?]
at com.mojang.datafixers.FunctionType$Instance.lambda$null$4(FunctionType.java:96) ~[minecraft_server.pre8.jar:?]
at com.mojang.datafixers.FunctionType$Instance$$Lambda$1510/1881320200.apply(Unknown Source) ~[?:?]
[22:41:29] [Server thread/WARN]: Can't keep up! Is the server overloaded? Running 315280ms or 6305 ticks behind
but it didn't crash. I have now logged in and can interact with the game. Until the next time this happens. I don't really want to leave the server with such a huge max tick time.
Flew around fast over built up areas.
[08:19:30] [Server thread/WARN]: Can't keep up! Is the server overloaded? Running 18355ms or 367 ticks behind
[08:19:51] [Server thread/WARN]: Can't keep up! Is the server overloaded? Running 5929ms or 118 ticks behind
[08:21:26] [Server thread/WARN]: Can't keep up! Is the server overloaded? Running 80632ms or 1612 ticks behind
[08:21:45] [Server thread/WARN]: Can't keep up! Is the server overloaded? Running 4187ms or 83 ticks behind
[08:22:36] [Server thread/WARN]: Can't keep up! Is the server overloaded? Running 35475ms or 709 ticks behind
[08:23:38] [Server thread/WARN]: Can't keep up! Is the server overloaded? Running 46850ms or 937 ticks behind
[08:24:05] [Server thread/WARN]: Can't keep up! Is the server overloaded? Running 2051ms or 41 ticks behind
[08:24:22] [Server thread/WARN]: Can't keep up! Is the server overloaded? Running 2057ms or 41 ticks behind
[08:25:11] [Server thread/WARN]: Can't keep up! Is the server overloaded? Running 2069ms or 41 ticks behind
[08:26:05] [Server thread/WARN]: Can't keep up! Is the server overloaded? Running 2067ms or 41 ticks behind
[08:27:04] [Server thread/WARN]: Can't keep up! Is the server overloaded? Running 44276ms or 885 ticks behind
but it didn't crash.
getting better, but it'd still have crashed if i hadn't upped the max ticks:
[16:43:48] [Server thread/WARN]: Can't keep up! Is the server overloaded? Running 19581ms or 391 ticks behind
pre10 still needs the max ticks set stupid high:
[08:56:44] [Server thread/WARN]: Can't keep up! Is the server overloaded? Running 20518ms or 410 ticks behind
functionally seems to be the same as:
https://bugs.mojang.com/browse/MC-132135
@sewan - we had to increase the max tick time to 10 minutes to stop crashes. The closest thing we can find to a pattern is that the conversion on some old chunks gets into a weird state sometimes but does eventually sort itself out? Not a great help
I'd like to applaud @Jessedward for her dedication in keeping the affected versions for this bug updated over three years.
This bit me yesterday - at least 3 strikes within 50 blocks, setting a large part of a village on fire. I've had the thunder on login but not noticed the lightning for whatever reason.
aye:
openjdk version "1.8.0_212"
OpenJDK Runtime Environment (build 1.8.0_212-8u212-b03-2~deb9u1-b03)
OpenJDK 64-Bit Server VM (build 25.212-b03, mixed mode)
If this is just works as designed - and given its been 3 months, I suspect so - can it be marked so?
Ugh sorry, just spotted
MC-145598You need to get a copy of open jdk 16, and install it too - its only in debian unstable at the moment so you'll need to do it manually, and set JAVA_HOME etc accordingly.
https://www.minecraft.net/en-us/article/caves---cliffs--part-i-out-today-java
solved.
Had to clear out old java options - in the launcher, under installations, edit the existing one, select more options, then reset the java options.
I suspect they were old ones imported over a long time now no longer valid for java16, but it was very obfuscated by what seemed like glx errors
yes, shockingly something changed in the last three years since it was reported.