Closed Consistent "can't keep up!" on VM box server.

[Essentiality]

Summary of the problem Consistent "can't keep up!" on VM box server.

Pack Version 1.2.0

What is the bug? The server seems to have the "Can't keep up!" message roughly every 14-20 seconds endlessly. We've attempted to unload specific chunks to see if it'd stop or change in consistency to no avail. We're also not convinced it is a hardware issue as I'll share specs below.

Mod & Version

Link to log file

Is it repeatable? Yes, it seems to always happen, even when changing JVM arguments, unloading chunks and trying new ones, etc.

Known Fix

 

[Essentiality]

The hardware running on the box is an Intel Xeon CPU E5-2670 @ 2.60 GHz with 10 GB of RAM and a 64-bit OS.
CPU Usage and a Speed Test are attached to this, as well as the log.

We also played around with the JVM arguments in an attempt to fix it. This is what we're set at currently, though I want to stress that all changes we made yielded no noticable results or change:


rem Do not touch
set MCVER=1.12.2
set JARFILE=minecraft_server.%MCVER%.jar
set LAUNCHWRAPPERVERSION=1.12
set LAUNCHWRAPPER=net\minecraft\launchwrapper\%LAUNCHWRAPPERVERSION%\launchwrapper-%LAUNCHWRAPPERVERSION%.jar
set FORGEJAR=FTBserver-1.12.2-14.23.1.2573.jar

rem can be changed by user
set MIN_RAM=4G
set MAX_RAM=4G
set PERMGEN_SIZE=256M
set JAVA_PARAMETERS=-XX:+UseParNewGC -XX:+CMSIncrementalPacing -XX:+CMSClassUnloadingEnabled -XXarallelGCThreads=2 -XX:MinHeapFreeRatio=5 -XX:MaxHeapFreeRatio=10


Any input on what might be the issue, please let me know! The server is still playable for everyone, I'd just like to make it better if possible.

Thanks.

 

[DrDurandal]

From your logs, it looks like you are getting messages when there are more than 2 players on the server.

What VM platform are you running on? I've seen similar behavior on AWS instances, which on paper should be good enough, but were only resolved by throwing more compute power at it.
https://forum.feed-the-beast.com/th...moving-to-resurrection-1-7.58961/#post-864851

 

[Essentiality]

It is running Hyper-V 2016 core and it is on a HP DL380P, with 32 (we're giving it 8) Xeon 2.8 Ghz processors with 128 GB of DDR3 RAM. It also has 8 200 GB server-grade SSDs on a RAID6.

We're going experiment with a Linux based system since this is using Win7 Pro 64-bit. We'll try MineOS here soon and I'll share the results. For science.

 

[DrDurandal]

Keep in mind single threaded performance > more cores for MC.

Look forward to your results.

 

[grandrolf]

It is running Hyper-V 2016 core

Host is forcing time-sync on the guest, either configure (should be able to do that), the service or the easiest way:

Turn off "Time synchronization" (integration services), on that VM - either using Hyper-V manager, vmm or PowerShell.

And configure time sync to run on the guest less often, and not from the host.

EDIT: question - have you turned on the raid controller instead of using storage spaces, or are you just having the raid controller in pass-through and using storage spaces? <-- has nothing to do with your issues with MC, just curious.