Clients losing connection with the server with the "timeout" reason

Problem 1:
Hey, I’m starting as a developer of a RolePlay server, I’ve been around for a month or so, and I’ve been noticing for about 2-3 weeks that there are quite a few users with server disconnections due to “Server-> Client connection timeout last seen (number) ms”, they get this message (excuse the quality, it is extracted from a video of a user that I send via ticket).

- Client error message:

- Server error message:
image

Problem 2:
I do not know in this case if I will have to see or not but users also complain of a notable desync (Bumps that some users see and others do not), I have solved this a bit with a server event that was spamming 24/7 , but they keep warning that there is desync.

Some images of graphics/tests performed:
- From txAdmin:

Profiler:

Network metrics:

Resource monitor:

This chart looks a bit odd - what hardware are you running the server on, and would it be possible to capture and upload a .etl file during such peak times?

I’m running the server on Windows, with the 3922 artifacts, and on an OVH machine, excuse me, but could you tell me how to do that with the .etl? I’m a newbie, excuse me :sweat_smile:

1 Like

Right now I have 90 people on the server, if you explain to me how I can do it or where I can find to do it I will.

Right now I watch it and I do it, how much do you recommend to have it recording?

Set ‘Tracing to file’ and record for around 20 seconds at most.

1 Like

Here you have it extracted with 90 people on the server.
2021-05-23_11-56-01_Micky.etl (57.3 MB)

And just today the players who are experiencing failures in the loading of ground textures etc …

Are any people having aforementioned issue around now?

Will be able to look later, only, sadly :confused: So if people aren’t having the issue yet maybe capture again when they do.

I have read the above message, players are experiencing texture loading glitches and low fps.

… but both don’t seem server-related directly. ‘Low FPS’ implies however that it should be possible to take a trace from an affected client as well, though!

I have read on the Internet and FiveM forums that this is a possible cause is that OVH DDoS mitigation may have something to do with it, could it be possible? I don’t know what could be more, the players tell me that the first moment they enter the server well, but after a while the failures begin.

Now I have noticed that there are times that this comes out to some users.

Oh that gcphone event looks particularly bad… heeeh.

Is it possible that if I have 148,387 messages from people in the gcphone database it is causing that error? And another possible conclusion from this is that the server’s Internet connection with some clients is slow.

I’d not be surprised if it turns out that gcphone does send a lot of message history even when not needed, large events are known to cause issues like these which is why they should be used sparingly.

Right there it gives huge FPS dips, between the dips and the falls of the people it is almost impossible to play and I for my part no longer know what to do since I have searched through forums and I can not find anything.

Fall issue I’m drawing the conclusion that it is something of the Internet between the server and the client, I do not know if I will have to increase the MB’s of the machine or what. And then another conclusion that I have drawn for those FPS drops is the high load of streams (maps, vehicles, clothes …)

1 Like

image
Could a possible cleanup in the message table of the gc phone database save all this?

Try it! Maybe make a backup first.