One cause of Ped Pool Crashes

[benzonyt]

A lot of servers spawn their players in the same place ex. What in RP is called the job center.

When a player spawns in NPC/PED’s start spawning outside, and that fine not really an issue, but when the player is TP’ed to there last saved location, these NPC/PED’s become a problem.

Since they are not released, they end up being stuck to the owner player that caused them to spawn! So they become buggy, can’t die don’t move, and just stand still, so the more that joins the higher the PED/NPC count.

IF every player on the server is returned to the spawned PED/NPC’s THEY are RELEASED, and able to move again, and move away from the area.

So what I do believes happening, is the NPC/PED’s fails to migrate, and get stuck to there owner entity the player.

1. Client (Prod/Canary) and FXServer version
   Both

2. What you expected to happen
   PED’s being released and moving away, later despawn like normal.

3. What actually happens
   PED’s freeze up, and bugs out.

4. Category of bug (eg. client, server, weapons, Peds, native)
   Server - Since this is not a thing in older FXServer versions NPC/PED’s

5. Reproduceable steps, preferably with example script(s)
   We tested out with a empty server, asked 6 players to join, they joined in spawned different places on the map, but just whent to the job center first, before selecting there char.

When they selected there char, more and more unresponsive PED’s pilled up outside.

What we then did was placing players different locations on the map, asked them to join again, the same issue happen, what I did then solved our mystery.

I tp’ed every player to me, when every player got in range of the NPC/PED’s they where released again.

We tested this 4 times, with the same result every single time.

What we did to migrate the issue for now until fixed, have 20 default spawn locations, when players are using our default ped, ClearAreaOfPeds and Vehicles run in a Thread with a while loop, when they change ped, this while loop will stop running, and stop Clearing area.

But this is a work around and a migration method not a fix.

[nta]

What server build is this on?

Can you induce this with, say, 3-5 players teleporting from one place to another on a clean server with no scripts other than, say, vMenu to allow teleporting?

If it doesn’t happen there, but does happen with your scripts, that’s an interesting correlation: I’m guessing some script is causing the client/server to not delete the peds when moving around from place to place.

[benzonyt]

if you want the old FXServer Version number we where using ill try and get that for you, but in that version the issue was not present at all.

[benzonyt]

Okay ill do my best to explain.

Artifact version is 2967, the issue is there with all newer versions of FXServer, and we have actually sticked to a older version for quite some time, since it was stable.

Ill try and test with vMenu not sure if i can repro it with vMenu since it’s abit of a wonkey issue.

Since it’s basicly char selection that causes the issue, since everyone spawn in as default ped andy, until they select a char, and then teleported out to last known location, so vMenu might be tough to repro with, since it just spawns you one of the random places and you are good to run around.

I do think the issue might be related to kashacters/spawnmanager in combination, we tryed to disable quite a few scripts to see if some of them would affect the issue, with out any luck.

[nta]

@benzonyt

Since it’s basicly char selection that causes the issue, since everyone spawn in as default ped andy, until they select a char, and then teleported out to last known location, so vMenu might be tough to repro with, since it just spawns you one of the random places and you are good to run around.

Either way, if you want it to be fixed you’re going to have to make it reproable… even sending a bundle of relevant server data files + empty database could help, but if the repro is ‘collect all the scripts yourself’ it’s a bit nah-implausible.

(fyi we’d a representative package of ESX resources and a database script for it for testing)

[benzonyt]

New discovery, so got logging on the amount of networked ped’s running, and auto clean up script.

But noticed out count is very low and fine nothing there.

BUT!!! The Clients have a higher count, and got none networked objects these peds are the one causing the issue, it seems like they aint cleaned up the way they should be in some wird way to be honest.