Oxygen-river-music after a few hours

FiveM Discussion OneSync reports
, , ,
1

Hello, below is the filled in bug report form. I hope you guys can be of assistance.

Client

Using canary? not relevant
Windows version: not relevant
System specifications: not relevant

Server

Operating system: Debian 9.8 Running Kernel version: 4.9.160
Artifact version: Latest Linux Build from 16th of febuary
IP address: 164.132.203.199
Resources: ESX Roleplay server running all the latest updates. Nothing particulary spectacular
System specifications: i7 4790k, 16GB Ram

Incident

Summary: Today we decided to have our second test running a OneSync 48 slot server for a prolonged period of time. It ran fine all the way from 5am till 3pm today but then suddenly started having people crash out upon joining. In the server console the message “couldn’t assign an object id for player” is spammed if someone is joining. We narrowed it down to just people joining because as soon as the queue was disabled and the server was full it stopped happening.
Expected behavior: People are able to join the server no problem
Actual behavior: People crash out during or after the loadingscreen due to the oxygen-river-music error. Meanwhile the server is continuously printing “couldn’t assign an object id for a player”
Steps to reproduce: Download latest linux build, install fivem-queue and have it run for a prolonged period of time with the server full and people joining and leaving as they wish. Notice people start crashing out.
Server/Client? Server
Files for repro (if any):
Error screenshot (if any):
.dmp files/report IDs:

Any additional info:
[16:18PM] NEW INFORMATION:
Seems to happen even with the queue turned off after a few hours

1 Like
2

It might be object IDs don’t get freed with players leaving in some cases.

Again, Linux does not provide the required diagnostic tools to figure out what is going on; if this were Windows a full server dump would easily be able to tell where object ID allocations go.

3

So basically transferring to Windows is the only option? Would it not be possible to get a updated build for Linux? Or atleast instructions on how to build from source.

4

That’s not the issue. The issue is:

Hence during the experimental phase of OneSync, Linux support is limited and these unfortunate crashes you get are close to impossible to fix :frowning: It sucks, I know, but there’s not much you or FiveM can do other than you switching to Windows.

Try and disable fivem-queue, see if that fixes it. If it does, try to maybe isolate within fivem-queue’s code where it fails. That’s my two cents.

5

Repost from #onesync
-> https://anonfiles.com/Z9JdH1v2b4/Procdump_rar
The requested dumps. You have two servers here ^

Another error appeared while connecting https://cdn.discordapp.com/attachments/538094621297999882/551783091379765287/unknown.png
I don’t know if it is useful or provide any information.

6

Alright, so since starting the topic we transferred to a Windows server (with pain in my heart)

The issue still seems to arise when running a few hours at 48 slots. Since we are now able to create a dump i have attached it to this post. It doesn’t seem to be related to fivem-queue as it also happens when this is disabled.

Dump: https://we.tl/t-FGiEGR6Xfh

1 Like
7

it might help: https://github.com/citizenfx/fivem/commit/004d299f722363239a728baacd6562d1a4910673

plus
image

8

In the last 24 hours, no one has encountered a similar error. :hugs:

9

yep seems fixed