Running 2 containers with one docker-compose file causes 'WARN mc-server-runner sub-process failed {"exitCode": -1}' -error

[taeraeyttaejae]

When running a single docker container, there is no problems whatsoever at least to this point, but when running 2 parallel server containers launched with docker-compose, either one container dies with 'WARN mc-server-runner sub-process failed {"exitCode": -1}' -error message.

VM has 2/16gb of allocated RAM(2 minimum, 16 max) and 10 xeon cores(yeah minecraft is single core but still)
paste from my docker-compose yml(excluding pwd:s etc):

https://pastebin.com/h6bafUXR
paste from 'docker logs mc1' container:
https://pastebin.com/92jLwYzP

That much i have gathered, that this probably has something to do with Docker memory handling. I am running MC server with default memory settings, using 1G min 1G max(?) of RAM.

Excerpt from container logs when starting the server:
[init] Running as uid=1000 gid=1000 with /data as 'drwxrwxr-x 7 1000 1000 4096 Jan 23 10:48 /data'

[init] Resolved version given LATEST into 1.18.1

[init] Resolving type given VANILLA

[init] server.properties already created, skipping

[init] Checking for JSON files.

[init] Setting initial memory to 1G and max to 1G

[init] Starting the Minecraft server...

Folder contents:
xxxx@ubuntu-server:~/minecraft$ ls -lah
total 24K
drwxr-xr-x 4 root root 4.0K Jan 24 14:35 .
drwxr-xr-x 9 xxxx xxxx 4.0K Jan 23 17:17 ..
-rw-r--r-- 1 xxxx xxxx 930 Jan 23 18:59 docker-compose.yml
drwxrwxr-x 7 xxxx xxxx 4.0K Jan 23 10:48 minecraft-server1
drwxrwxr-x 7 xxxx xxxx 4.0K Jan 23 12:12 minecraft-server2
-rwxr-xr-x 1 root root 304 Jan 17 21:55 start_mc.sh

minecraft-server1 and 2 are folders that are mounted on containers' /data, they should be also backed up from here in the undeniably unsure future when my servers actually work...

Any ideas where to go next? I was burning midnight oil with this last night already.
What kind of additional details should i provide?

[itzg]

Your mc2 port mapping should be 25566:25565; however that won't be the immediate cause. Can you provide the exact docker compose command you're using to bring them both up?

[taeraeyttaejae]

docker-compose -f docker-compose.yml up -d

[taeraeyttaejae]

and yes, i have corrected the port mappings after that paste. It seems that network is not the problem here but java in some...javaish way.

[taeraeyttaejae]

could the server.properties -files in /home/xxxx/minecraft/minecraft-server1(and2) -folders collide somehow with the environment variables here, as at least MOTD messages seem to come from those files?

[shotah]

@taeraeyttaejae were you able to just run 1 of the servers? The exit code seems interesting because it's still in world gen mode. So you've passed all the setup scripts. So is it trying to write the world data files and failing? I guess it'd be nice to verify it is something with 2 containers vs. 1 container.

[taeraeyttaejae]

I tried to run 2 separate containers, but i ran into same problem with same error message.

[shotah]

just tried two and one and its giving a corrupted world file in all cases.

[taeraeyttaejae]

ok this is good to hear, that someone else is having the same problem.

Any ideas how i should run 2 LAN servers in relatively simple setup(proxmox ve, ubuntu 20.04 server)?

[shotah]

I think we need to wait for Itzg to reply. He may have more insight. I am unable to stand up 1 container at the moment.

[taeraeyttaejae]

Yeah I think this is an issue many people have ran into - at least my kids like to have separate survival and creative servers to choose from.

[itzg]

I'm thinking it's the health check timing out since two containers both building two world datas from scratch gets pretty intensive. I adjusted the compose file to use relative paths and default rcon settings, but increased the health check allowance and they eventually started:

version: "3"
services:
  mc1:
    image: itzg/minecraft-server
    container_name: mc1
    ports:
      - 25565:25565
    environment:
      MOTD: "Survival server"
      EULA: "TRUE"
      ONLINE_MODE: "FALSE"
      restart: always
    volumes:
    - ./minecraft-server1:/data
    healthcheck:
      interval: 10s
      timeout: 30s
      retries: 24
      start_period: 30s
 
  mc2:
    image: itzg/minecraft-server
    container_name: mc2
    ports:
      - 25566:25565
    environment:
      MOTD: "Creative server"
      EULA: "TRUE"
      ONLINE_MODE: "FALSE"
      restart: always
    volumes:
    - ./minecraft-server2:/data
    healthcheck:
      interval: 10s
      timeout: 30s
      retries: 24
      start_period: 30s

[shotah]

I tried adding depends_on as well. My local version (Which I normally don't use) isn't honoring it.

mc2:
     ...
     depends_on:
       - "mc1"
     ...

[itzg]

I tried that too 😀 and also observed it not getting honored. That was unexpected, so I switched gears to the extended health check.

[taeraeyttaejae]

I had to change
version: "3"
to
version: "3.7"
otherwise i got
ERROR: The Compose file './docker-compose.yml' is invalid because: services.mc1.healthcheck value Additional properties are not allowed ('start_period' was unexpected) services.mc2.healthcheck value Additional properties are not allowed ('start_period' was unexpected)

Also, probably ports are still wrong or something, "mc2" does not allow connecting to it yet.

I am not 100% sure about this yet, i will mark this solved when i can get both servers rollin'

STILL, thanks a bunch man, couldn't have came up with the health check!

[taeraeyttaejae]

ok, i tried to run my separate "mc2" and it does not work alone, there is something else amiss with it still!

If there is some way to support your project, i would be glad to drop couple euros somewhere :)

Thank you for your efforts, and for this effort.

[itzg]

Try moving away the minecraft-server2 directory and try the mc2 by itself again. If that works individually, then mc1 by itself, finally should be better at running together since the world gen would be prepared.

Since you asked: https://www.buymeacoffee.com/itzg

[taeraeyttaejae]

I think i found the root cause for all my troubles, it was memory -related most probably.
https://imgur.com/a/uosyMwr <-- screencap of my Proxmox memory settings. I changed minimum memory from 2048 to 4096 and it looks like there are no more problems with running 2 servers. I have experience over 15 minutes of running but will update here if things change.

It seems this was not image-related, docker-related or OS -related issue, but underlying virtual environment -related problem. Thank you all for your support in this issue!