Exception on org.bukkit.event.inventory.InventoryClickEvent.getCurrentItem

[LemADEC]

Server Log:

[14:11:38] [Server thread/ERROR] [Minecraft/]: Could not pass event InventoryClickEvent to WorldGuard v6.1.1-SNAPSHOT.1616-c9eaec0
org.bukkit.event.EventException
    at org.bukkit.plugin.java.JavaPluginLoader$1.execute(JavaPluginLoader.java:338) ~[JavaPluginLoader$1.class:1.7.10-1614.54]
    at org.bukkit.plugin.RegisteredListener.callEvent(RegisteredListener.java:62) ~[RegisteredListener.class:1.7.10-1614.54]
    at org.bukkit.plugin.SimplePluginManager.fireEvent(SimplePluginManager.java:507) [SimplePluginManager.class:1.7.10-1614.54]
    at org.bukkit.plugin.SimplePluginManager.callEvent(SimplePluginManager.java:492) [SimplePluginManager.class:1.7.10-1614.54]
    at net.minecraft.network.NetHandlerPlayServer.func_147351_a(NetHandlerPlayServer.java:2028) [nh.class:?]
    at net.minecraft.network.play.client.C0EPacketClickWindow.func_148833_a(SourceFile:32) [ix.class:?]
    at net.minecraft.network.play.client.C0EPacketClickWindow.func_148833_a(SourceFile:10) [ix.class:?]
    at net.minecraft.network.NetworkManager.func_74428_b(NetworkManager.java:245) [ej.class:?]
    at net.minecraft.network.NetworkSystem.func_151269_c(NetworkSystem.java:173) [nc.class:?]
    at net.minecraft.server.MinecraftServer.func_71190_q(MinecraftServer.java:1013) [MinecraftServer.class:?]
    at net.minecraft.server.dedicated.DedicatedServer.func_71190_q(DedicatedServer.java:431) [lt.class:?]
    at net.minecraft.server.MinecraftServer.func_71217_p(MinecraftServer.java:831) [MinecraftServer.class:?]
    at net.minecraft.server.MinecraftServer.run(MinecraftServer.java:683) [MinecraftServer.class:?]
    at java.lang.Thread.run(Thread.java:745) [?:1.8.0_77]
Caused by: java.lang.NullPointerException
    at org.bukkit.craftbukkit.v1_7_R4.inventory.CraftInventoryView.getItem(CraftInventoryView.java:64) ~[CraftInventoryView.class:1.7.10-1614.54]
    at org.bukkit.event.inventory.InventoryClickEvent.getCurrentItem(InventoryClickEvent.java:102) ~[InventoryClickEvent.class:1.7.10-1614.54]
    at com.sk89q.worldguard.bukkit.listener.BlacklistListener.onInventoryClick(BlacklistListener.java:264) ~[?:?]
    at sun.reflect.GeneratedMethodAccessor60.invoke(Unknown Source) ~[?:?]
    at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43) ~[?:1.8.0_77]
    at java.lang.reflect.Method.invoke(Method.java:498) ~[?:1.8.0_77]
    at org.bukkit.plugin.java.JavaPluginLoader$1.execute(JavaPluginLoader.java:334) ~[JavaPluginLoader$1.class:1.7.10-1614.54]
    ... 13 more

Explanation of issue:
Console get spammed with exception from all plugins trying to call getCurrentItem().

How to recreate this issue:
Not sure, it happens on our server, about once per hour.

Modpack Name: (Only if public)
WorldWarMinecraft (official)

Mods Installed:
see modpack page on Technic, version 8.1.0

Plugins Installed:
not relevant

Thermos/Forge Version:
Thermos-1.7.10-1614-54

[Time6628]

Could you try build 55?

[LemADEC]

I would prefer not try a bleeding version on my public server :)

[Time6628]

@LemADEC It works just fine, I use it on all 20 of my servers 😸

[sameer]

It was marked bleeding because I was unable to test it at the time, it is essentially a beta.

[djoveryde]

Build 55 is safe. I use it 👍

[LemADEC]

25 hours after installing Build 55, we've already experienced 3 worlds corruption: 1 chunk reset and 2 world rolling back.
Compared to previous version, over 4+ weeks, we had only 1 chunk reset.
As a preventive measure, I'm rolling back to build 54.

Regarding the tile entities not saved, it comes down to case 1:
1- player moves things from his ender chest to a block (ME drive)
2- player moves to another dimension => dimension unloads (no logs available to confirm)
3- about 5+ minutes later, player moves back to that dimension
At that point, player still has the modified inventory but blocks are 'older' than last change.

case 2:
1- player moves things from his inventory to a block (Thaumcraft arcane workshop for charging)
2- player moves to another dimension
3- server restarts
4- about 15+ minutes later, player moves back to that dimension
At that point, player still has the modified inventory but blocks are 'older' than last change.
There's no specific logs at step 2 nor during server shutdown nor step 4.

[sameer]

Tu peux nous expliquer un peut plus svp?
"3 worlds corruption:" c'est a dire que le monde avait ete perdu completement?

Je ne crois pas qu'il y a une grande difference entre 54 et 55, cependant, c'est possible.

Ceci montre le difference entre les deux: 54...55

[LemADEC]

case 1 and case 2 looks like a specific chunk (region?) wasn't saved and when requested again, server pulled from disk an 'old' version of it.

case 3 is:
1- player sets home at his base at 2016-04-22 10:39:58, ME system is present
2- player wanders around mining
3- player comes back to base at 2016-04-22 10:45:27 (~5m29s later)
=> chunk containing the ME system is reset
There was no server restart during that time.
No specific entries in logs, just a player connecting at 10:45:00 into the same dimension.

Entre 54 et 55, je vois des changements sur la gestion des chunks (hashs et checkChunksExist en particulier).
Je n'ai pas vu d'erreur type memory leak dans les logs, mais nous avons probablement une corruption mémoire qui traîne: la VM se bloque après ~2h30 d'uptime. Lorsqu'elle est bloquée, les debugger n'arrivent pas à se connecter à la VM, y compris par les commandes de bas niveau :(.

[Time6628]

What is your autosave interval set to? This happens sometimes when the chunk doesn't save before it is unloaded.

[LemADEC]

It's up to the server to save chunks before releasing them from memory. Autosave is only there in case the server have an hard crash, which is not the case here.

[sameer]

@LemADEC Utilises-tu ovh? C'est possible qu'il y a un probleme avec ton serveur. Tu as deja vu les logs "S.M.A.R.T." pour tes disques durs?

[LemADEC]

oui, c'est du OVH. Par contre, c'est du SSD redundant, ce serait vraiment pas de bol que cela soit ça.
Reste à me souvenir comment lire le status des disques...

[sameer]

M'aussi avait SSD 2x480 RAID, mais ces disques ont un tendance de rater avant que des HDDs

[LemADEC]

le RAID est sain, les rapports SMART sont Passed.

[sameer]

Hmm, c'est tres etrange, je vais verifier le code entre 54 et 55 ce weekend, mais aussi j'ai besoin de publier build 56 😕