Large World Anchor Logfile causing crashes?

[Linkeleo]

Title: Large World Anchor Logfile causing crashes?

Version: 3.1.2

OS: Linux / CentOS 6 x64

Java Version: 1.7.0_04

Description of Problem:

My server has been stopping with no errors or even overload warnings recently, about every 12-24 hours or so. After doing a bit of looking around, I came across a "dimensional-anchors.log" file in the server directory that was almost 700MBs. I found that, if I attempted to open this file with gedit while the server was running, it would quickly crash the server in the exact same manner as the seemingly random shutdowns. (gedit would also crash at around the same time).

I've deleted the log file and changed the "chunkloader.logFileName=" value in config/immibis.cfg to blank. Everything seems to be running OK at the moment.

A very frustrating issue, especially since user's inventories were corrupted by one of the crashes. Apologies if this has been posted before, I looked but didn't come across anything.

Again, there were no errors at the time of the crashes, the terminal it was running in simply went back to standard prompt without any warning.

Error Messages:

None

Error Log:

[Skuli]

My dimensional anchors log file is 2gb. :(

[Phuriousgeorge]

If you go into your config folder, open the immbis.cfg (I think that's how it's spelled-sorry, not at my home PC at the moment).

After you open it, look for the filename that is growing entirely too large and remove all characters after the = sign. The server will no longer log to that file. Not needed anyway and imo should be turned off by default.

[Linkeleo]

Well it seems this wasn't the issue (Although a 700MB log file after only one day is an issue by itself) causing the crashes, as the same thing happened earlier today. No warnings, errors, information or anything, the server just stops. Any idea what could be causing that or where I could find more info? I've ran a tekkit server on this server for several months and this is the first time I've had this problem.

The server has 2x Dual Core 3GHz processors, 15K RPM SAS Drives, and tekkit has 3GBs of ram dedicated to it; there isn't much else running on the machine atm either. I have WorldGuard, WorldEdit, WorldBorder, Dynmap, Multiverse, Essentials, PermissionsBukkit, and ChopTree plugins installed.

Thank you for any help.

[Phuriousgeorge]

Are you running MultiWorld or MultiVerse? I didn't have significant issues until I started fooling with MultiWorld, then everything seemed to take a crap. Everything from plugins to random lock-ups and restarts. Never had this issue until 3.1.2, but can't identify a specific cause either.

[Linkeleo]

I'm using MultiVerse Version 2.4-b527. Along with Multiverse-Portals and Multiverse-NetherPortals.

I'm pretty sure I didn't have this issue before 3.1.2 also.

[Linkeleo]

Sorry for the bump, but has anyone experienced this kind of thing or know of a way to get more information on whats happening?

I've been trying uninstalling plugins, but I don't know how to reproduce the issue other than waiting 12-24 hours for it to crash again, usually dumping everyone's inventories

Thanks

[Phuriousgeorge]

How's your resource useage? CPU/RAM? I've ran into similar issues with the server just dropping and I never experienced it until I installed multiple world plugins. Things have gotten better, but I can't specifically point out what I've done to fix it as when it takes a crap, there's no error code. It just goes *poof*.

Good thing is it's on my test server and not my live.

[Linkeleo]

According to top, CPU is usually around 20-35%.

The "GC" command (in essentials I believe) gives a TPS of 20,

Max Memory of 2,730 MB, Allocated: 1,965MB, and Free: 1,455MB

I've run this server with multiverse for several months and this is the first time I've had this issue, seems to have started with 3.1.2 (upgrade from 3.0.4).

[xSirStoner]

Did you still find a fix for this? I have the same issues.

I fixed it so far by disabling the nether. It seems to be caused by something in the nether (presumably nether ores). Please let me know

[Linkeleo]

Can't say I was ever able to identify the cause, I set up several servers with different things disabled and wasn't able to pinpoint the issue. Never thought to disable the nether tho...

What I did end up doing tho, that stopped the crashes, was implement an automatic server restart every so many hours. Since that "fixes" it, I am guessing it has something to do with memory, but again, I was never able to figure out what exactly.

Sorry I couldn't provide you with a better answer :-/

[Kenji_03]

The "dimensional-anchors.log" file is exactly what it's extention hints at a log of every time the Dimensional Anchor checks to see if a player is on the server. The reason the file gets so big is that after just a few seconds it will log 6kb of entries! It will create absolutely enormous files in very short order (1mb every 2.8 minutes, and 1gb every 2.25 days).

The problem is that this constant log is pretty much useless, as I can't think of any reason I'd want to be able to open up the file and see the exact second one of the worlds was loaded (which is 40 times a second -- fyi)...

I'm currently trying to find a way to disable this pointless file, and if you also think it's ridiculous having this file bog up our hard-drives, contact the mod creator here (remember to be polite, he does make these mods for free and they are rather labor intensive)