Mekanism

Mekanism

111M Downloads

Server Hang Issue with TileEntityMekanism.insertEnergy

windy664 opened this issue ยท 7 comments

commented

Issue description

While running a Minecraft server with the Mekanism mod installed, I encountered a server crash. The crash report indicates that the server hang was detected due to a single server tick taking 60.00 seconds, whereas it should be a maximum of 0.05 seconds. The issue seems to stem from the TileEntityMekanism.insertEnergy method in the Mekanism mod. The expected outcome would be for the server to run smoothly without any extended tick durations.

Steps to reproduce

1.Set up a server with the specified versions of Minecraft, Forge, and Mekanism.
2.Observe the server tick duration and potential server crash with the mentioned issue.

Minecraft version

1.20.1 (Not yet released, manually compiled)

Forge version

1.20.1-47.1.3

Mekanism version

Current Dev Branch (Manually Compiled)

Other relevant versions

No response

If a (crash)log is relevant for this issue, link it here: (It's almost always relevant)

https://gist.github.com/windy664/3ebffb606bea5b5353a314d04beb80b1

commented

Last version of Mekanism is v1.19.2-10.3.9.13

Are you sure, you are running forge/mc 1.19.2 ?

commented

He said that he manually compiled 1.20.1

commented

The server watchdog tends to not be the best indicator of what is necessarily causing the lag as it just includes in the stack trace what was going on when it finally hit the limit. I would recommend trying to set max-tick-time to -1 in the server.properties to disable the watchdog, and then using a profiler like spark to create a profile to get a better idea of what is using up all the tick time.

commented

Hello.

I couldn't find any bugs in Mekanism that indicate importing energy.
Testet Client and Server side.

Apart from that, I couldn't find any other bugs or shortcomings that affect play or compatibility with other mods.

2023-08-07_04 32 52

Screenshot 2023-08-07 032909

Screenshot 2023-08-07 032949

Screenshot 2023-08-07 033046

Screenshot 2023-08-07 033541

Screenshot 2023-08-07 035046

commented

Server side:
debug.log
latest.log

Client side:
debug.log
latest.log

commented

a watchdog crash doesnt indicate what took so long, only what was running when it decided enough is enough.

commented

The server watchdog tends to not be the best indicator of what is necessarily causing the lag as it just includes in the stack trace what was going on when it finally hit the limit. I would recommend trying to set max-tick-time to -1 in the server.properties to disable the watchdog, and then using a profiler like spark to create a profile to get a better idea of what is using up all the tick time.

When I update to the latest version, this problem does not appear .