Random super high ping lag spikes with solid 20 TPS. No idea what is causing them.
So intermittently we will get a huge lag spike on the server. Most players pings will shoot up to 300+ and sometimes even 1000+. This lasts about 30 seconds and slowly goes down over time. We have rubberbanding issues. All while the TPS is locked at a solid 20.
Some Info
* Server CPU: 3 Threads of 7950x
* Ram: 8GB
* Players: Usually around 3-8 (issue gets worse with more players online)
* Server Version: Purpur 1.21.1-2329-803bf62 (MC: 1.21.1)*
Troubleshooting I've Tried
* Lowering render distance
* Continuously pinging the server while the server is lagging in-game with command prompt. Command prompt stays at a solid 50ms ping for me while in-game it shoots up to 500+.
* We tried moving our server to a different location with the same host
* Tried a different host with the same CPU (7950x). Same issues.
Data Packs
* https://modrinth.com/datapack/refined-advancements
* https://modrinth.com/datapack/banner-flags
* https://modrinth.com/datapack/gardeners-dream
* https://modrinth.com/datapack/better-hanging-signs
* https://modrinth.com/datapack/mastercutter
* https://modrinth.com/datapack/rotten-flesh-to-leather
* https://modrinth.com/datapack/hopo-better-underwater-ruins
* https://modrinth.com/datapack/hopo-better-mineshaft
* https://modrinth.com/datapack/tidal-towns
* https://modrinth.com/datapack/dungeons-and-taverns-stronghold-overhaul
Plugins
* See screenshots below
Spark
https://spark.lucko.me/XiEf1Eu87f
Any Ideas?
I can't figure out what is going on. Is the only solution to just try to start testing if removing plugins one by one solves the issue? Are there any troubleshooting steps that I am missing or does anyone have any ideas what's going on?
29 Replies
Thanks for asking your question!
Make sure to provide as much helpful information as possible such as logs/what you tried and what your exact issue is
Make sure to mark solved when issue is solved!!!
/close
!close
!solved
!answered
Requested by faultyfunctions#0
could it be the entities loading in
ie when someone loads a chunk with like 1k entities
Hmmm I will take a look into that. Do you know how I can go from region number -> coords in-game?
x 16
/spark profiler start --thread *
Spark Profiler
Spark can be used to see why your server or client is lagging.
Users can share a spark report via
/spark profiler start --timeout 300
which creates a report after 5 minutes. The spark report is useful in debugging why your server is lagging, so ideally you should always provide one while asking for support.
Download | Website | DocsAdmincraft Canned Responses
make report with that
this sounds like netty thread getting overloaded
I don't have any players on right now. Should I wait until more people are on stressing the server? Can I simulate more players being on by raising the render distance?
Here is the spark with no much lag with a high render and simulation distance (I didn't get high ping during this because there was only me on):
https://spark.lucko.me/StDPneDiyU
I'll generate a new one once more players show up!
yeah wait till the issue appears
thats not... good
probably bluemap is putting too big of a load on your server if you only have 2 threads available
BlueMap
Configuration
BlueMap is a Minecraft mapping tool that creates 3D models of your Minecraft worlds and displays them in a web viewer.
We usually have 3 threads available. Itβs only 2 right now because we switched hosts to test if it was just our hosts network issues (which it wasnβt). Lag with 2 threads in virtually the same with lag with 3 threads. I mean it could still be blue map tho
I will test with the player render limit, thanks! And Iβll post an updated spark profiler when more players are on.
to be completely honest, I hate spark's all thread profiler. They don't have any filter for showing threads that aren't just sleeping, which makes finding the thread using the CPU really hard...
I've contacted our server host and they temporarily increase our thread count to 5 and our ram to 12GB. We'll see if that fixes it. I still think there might be some sort of underlying issue, but if that fixes it then I can't really complain haha
Upgrading our server seems to have solved the issue. Maybe with all the plugins we had, the server just wasn't able to keep up? Still weird that it manifested as a ping issues instead of a TPS issue?
Netty threads being overloaded would cause it
What would cause that you think?
Which would also explain why more threads helped
Well, theres many possible reasons
haha fair
Well here is the final spark report:
https://spark.lucko.me/waLPdUWqqD
spark
spark is a performance profiler for Minecraft clients, servers, and proxies.
In case there is something I missed
Heyo it seems like this is back for me? Here is the most recent spark report: https://spark.lucko.me/dMXo487ysy
spark
spark is a performance profiler for Minecraft clients, servers, and proxies.
We upgraded the server and it went away for a while but now two of our players are experiencing wicked high ping and disconnections. Any thoughts?
@ProGamingDk how to check netty thread performance to see if its getting overloaded π
Oh dang I used
--threads
instead of --thread
π© I'll run it againspark
spark is a performance profiler for Minecraft clients, servers, and proxies.
@ProGamingDk would you be able to take a look when you get a chance. I'm not sure what I'm looking for. π
I think I may have figured out the issue? Setting
MapPacketSendingRateLimit
in the ImageFrame plugin to 1
from -1
seems to have maybe fixed it? I think it was maybe sending too many map packets to players which was causing high ping and time outs on poor connections. Only time will tell though.