Forums Archived

This forum has been archived. No new posts can be made and no new users can sign up. It remains here for reference only.

Find the new forums here

7 Days to Die - High CPU usage for AMP_Linux_x84_64

  • Last Post 31 December 2021
AbhorrentJoel posted this 28 December 2021

OS Name/Version: Ubuntu 20.04.3 LTS

Product Name/Version: AMP v2.3.0.8

Problem Description: Sometimes, after the 7 Days to Die server is executed, the CPU usage is very high for AMPLinuxx86_x64.

The CPU reaches over 300% (using almost all 4 cores) alone for the AMPLinuxx86_x64 executable while the server itself may only be using 20-50% under no or light load. Even after the server is stopped, the usage remains very high until the instance is stopped.

I am using Docker.

Steps to reproduce:

  • Install AMP (enable Docker for instances)
  • Create a new instance for 7 Days to Die
  • (Maybe optional) Set world size 10240, RWG
  • Launch the server and monitor resource usage

Actions taken to resolve so far:

  • Tried to create a new instance without Docker - not been able to reproduce with that so far.

I should note that the high CPU usage does not always seem to happen, and it seems to be incosistent when attempting to reproduce. I apologise that I am lacking more details, but it has happened a few times and I am not sure what the exact trigger is so I do not know what to do next. I have already recently reinstalled AMP latest version, and when doing so I created a new AMP user after deleting the old (so it is basically a fresh install).

Order By: Standard | Newest | Votes
AbhorrentJoel posted this 30 December 2021

I have enabled debugging on this instance so I am going to keep an eye out on when the issue presents itself.

AbhorrentJoel posted this 31 December 2021

I've created a new instance and downloaded 7 Days to Die again. I went to start the server and it wouldn't generate a world with my settings. I aborted and changed the settings again, which wouldn't work so I aborted. Then the high CPU usage for the AMP_Linux_x86_64 started.

There is nothing immediately outstanding from the debug logs I can see. However, netstat still shows that RCON is "CLOSE_WAIT", which means RCON is probably still running and exhausting CPU cycles. This would imply that aborting the server does not stop RCON. And each time the server is aborted, AMP is still keeping the RCON open and it stacks, using more CPU each time.

Just to note, I don't think the server not launching is anything to do with AMP. I think the server itself doesn't like the options I've given it.