Leaking worlds

[Triblade_8472]

Title: Leaking worlds

Version: 0.5.1

OS: Windows Server 2012

Java Version: 1.7 Update 13

Description of Problem:

After upgrading from 0.5.7 to 0.6.1 my 'worlds' seems to have sprung a leak.

I've deleten the folders 'config', 'coremods' and 'mods' before copying the newer files from the newest server download. Also I've overwritten the .jar files.

0.5.7 worked perfectly on my Windows Server 2012 system.

And somehow my world suddenly saves itself every minute or so. This working as intended?

(Also, somehow in this bug report I can only chose 0.5.1 as the Tekkit Lite version...)

Error Messages:

Detected leaking worlds in memory.

Error Log:

2013-03-19 09:41:11 [iNFO] [sTDOUT] ComputerCraft: RedPowerWorld and methods located.

2013-03-19 09:41:52 [iNFO] [ForgeModLoader] Unloading dimension 1

2013-03-19 09:41:52 [iNFO] [ForgeModLoader] Unloading dimension -23

2013-03-19 09:41:55 [iNFO] [sTDOUT] saving

2013-03-19 09:42:00 [sEVERE] [ForgeModLoader] Detected leaking worlds in memory. There are 4 worlds that appear to be persisting. A mod is likely caching the world incorrectly



2013-03-19 09:42:00 [sEVERE] [ForgeModLoader] The world b21e4c (Tri-Tekkit) has leaked.



2013-03-19 09:42:00 [sEVERE] [ForgeModLoader] The world d8afa2 (Tri-Tekkit) has leaked.



2013-03-19 09:42:40 [iNFO] [sTDOUT] saving

2013-03-19 09:43:25 [iNFO] [sTDOUT] saving

2013-03-19 09:44:10 [iNFO] [sTDOUT] saving

[okamikk]

you need new worlds, mod updates and ~ID changes have broken your worlds to the point that they just crash.

[Gazaka]

Interesting because I get this error on a freshly created world. 4 Worlds leaking on a new start sounds like something up with the mods from the get go. Unless it's something else, would be nice to hear if any solutions out there.

[Malawi]

I get the same error both on 0.6.0 and 0.6.1

I assume that this is a well known error, and that someone is working on it. Was actually a bit surprised to see it in 0.6.1 since it's a recommended build.

[okamikk]

... you may need to install an earlier version of java, i've heard that helps sort some problems out, no idea why.

[teh_meetch]

It seems to be an issue with the redstone ether. The error log mentions dimension 1 and -23. In the Tekkit_Lite_Server_0.6.1\world directory there are three folders named DIM1, DIM-1, and DIM-23, all with redstone ether folders in them. Would disabling the Wireless Redstone mod fix it?

[Elias Aarekol]

Im having this problem for 0.6.5, and i cant fix it

Anny help`???

[wangmauler]

I'm having same issue on two freshly formated win 7 PCs, both with the newest tekkit lite, MC, and tekkit launcher.

Both machines say:

[sEVERE] [ForgeModLoader] The world 3096f45c (world) has leaked.

[sEVERE] [ForgeModLoader] The world 46aff06c (world) has leaked.

[CleverBullet]

I feel like this problem has gotten worse with 0.65.With 0.61 i only had to reboot my server every couple of days, now i have to reboot it every 3 hours.

It's a standard unmodified Tekkit Lite distribution. It runs with max heap size of 1.5GB.

[Triblade_8472]

I feel like this problem has gotten worse with 0.65.With 0.61 i only had to reboot my server every couple of days, now i have to reboot it every 3 hours.

It's a standard unmodified Tekkit Lite distribution. It runs with max heap size of 1.5GB.

First of all, thanks for trying to help all of you!!

I read somewhere this had to do with the 'leak detection thingy', but I can't seem to find this post anymore :s

Now I have the same problem as CleverBullet. The server worked good, more or less, but since the upgrade to 0.6.5 I have to restart the server every day/couple of hours as well... It crashes every day with -Xmx and -Xms options set to 1024M. Without these options the server doesn't seem to crash anymore, but just lags very badly to poor memory usage.

(And using 64-bit java instead of 32-bit java doesn't seem to make a difference)

For crash log lovers the first piece of the log here:

---- Minecraft Crash Report ----

// Why is it breaking 

 

Time: 27-4-13 1:44

Description: Exception in server tick loop

 

java.lang.OutOfMemoryError: Java heap space

    at java.util.HashMap.<init>(Unknown Source)

    at java.util.HashMap.<init>(Unknown Source)

    at bq.<init>(SourceFile:12)

    at buildcraft.transport.PipeTransportLiquids.writeToNBT(PipeTransportLiquids.java:329)

    at buildcraft.transport.Pipe.writeToNBT(Pipe.java:194)

    at buildcraft.transport.TileGenericPipe.b(TileGenericPipe.java:104)

    at aam.a(AnvilChunkLoader.java:311)

    at aam.a(AnvilChunkLoader.java:127)

    at im.b(ChunkProviderServer.java:232)

    at im.a(ChunkProviderServer.java:284)

    at in.a(WorldServer.java:844)

    at net.minecraft.server.MinecraftServer.a(MinecraftServer.java:373)

    at net.minecraft.server.MinecraftServer.q(MinecraftServer.java:605)

    at net.minecraft.server.MinecraftServer.run(MinecraftServer.java:497)

    at fy.run(SourceFile:849)

 

 

A detailed walkthrough of the error, its code path and all known details is as follows:

---------------------------------------------------------------------------------------

 

-- System Details --

Details:

    Minecraft Version: 1.4.7

    Operating System: Windows Server 2012 (amd64) version 6.2

    Java Version: 1.7.0_21, Oracle Corporation

    Java VM Version: Java HotSpot 64-Bit Server VM (mixed mode), Oracle Corporation

    Memory: 14028456 bytes (13 MB) / 1041235968 bytes (993 MB) up to 1041235968 bytes (993 MB)

    JVM Flags: 2 total; -Xmx1024M -Xms1024M

    AABB Pool Size: 1494 (83664 bytes; 0 MB) allocated, 1409 (78904 bytes; 0 MB) used

    Suspicious classes: FML and Forge are installed

    IntCache: cache: 0, tcache: 0, allocated: 3, tallocated: 63

    FML: MCP v7.26a FML v4.7.35.556 Minecraft Forge 6.6.2.534 75 mods loaded, 75 mods active

[phishforever420]

I have the same errors as well. New worlds as well as old ones spring leaks within hours of starting up the server. Getting the same :

[sEVERE] [ForgeModLoader] The world 3096f45c (world) has leaked.

[sEVERE] [ForgeModLoader] The world 46aff06c (world) has leaked.

Has anybody made any progress to tracking down the root cause of this issue?