#dedicated-servers
1 messages Β· Page 138 of 1
I loved it while on phone support when people told me that so I could drop the customer service and just talk jargon
Haha, same here
Ive setup quite a few SF servers over the last few years, but this one just does not want to connect
Turned it off as a "im not asking anymore, just work"
Lmao
Literally nothing, ive hosted them at OVH and theyve worked, which is whats totally baffling me
Yup, i can see the server, and its running, i can access the admin, setup a new game, everything works fine, its literally just on join
I hit the remote restart about 30 minutes ago, as a "Now im definately not asking" but alas
I say dumb question because I called my boss after failing to get a system to connect. After 20 minutes we determined I forgot to start the service on the server... It's been his first question every time since. It's been 2 years, and I still haven't lived that down.
Lol. Right?
I even tried different installs, so one from Epic, one from steam, nothing
Yeah, thats whats confusing, same on both Epic and Steam
Both with the same error, albeit different client ID's
Yup, online in both
And you can ping those ports on the server from your machine?
Trying to think of what could be causing this. Someone else is likely to come in here and figure it out immediately.
I had a similar issue where I could log into the admin console just find, but it would crash when trying to join. It ended up being caused by port 8888 being closed.
It will start at 8888 and keep going up until it finds an unused one. 8888 is also used as an http back up port. Can you open 8888 through 8950?
Already opened, just getting my collaegue to give it a try from his side
Colleague literally getting the exact same error
Open in your local firewall?
I'm stumped because when I had this issue it ended up being that damn port.
Im glad your stumped too because it honestly has been paining me for days lol
[2025.05.30-13.01.33:881][736]LogGame: Warning: Set Online State For Player : true - BP_PlayerState_C_2147472138
[2025.05.30-13.01.33:881][736]LogGame: Warning: Could not process player whose OnlineState changed: local user was invalid```
Logs from my collageue trying to join
I am the collageue, just assigned the additional ports of 8888-8950 to the container as I don't think those were in the original docs we looked at
Yeah, 8950 was arbitrary.
And we're in!
It counts up up to 251 ports
8888 wasnt in those docs we found, argh, but thats done it!
We're running it in Docker so we need to assign all of the potentially required ports ahead of time
Ours mentioned 7777 - 7850
Yeah, it's not really mentioned except as a footnote
Yes⦠unless the (un)reliable port problem gets fixed, I think VPN tunnels without forwarding the port across the router is going to be the best way to keep the server from crapping itself due to random port traffic.
Tailscale is nice for making wireguard seamless across dynamic addresses.
Did not know that. I'm also Docker stupid.
We both work for quite a large IT firm, docker is our goto π
Half of my homelab is powered by docker and spite
I set my ExternalReliablePort to a 5 digit port, then forwarded that port to 8888. ReliablePort never worked, but the External one did.
Combined with the Ubiquiti router's firewall blocking most port scanners, plus my growing list of rules from security whack-a-mole, my server has been up for almost two weeks.
I really need to learn Docker.
Mine is powered by Proxmox and swearing.
Proxmox is fine, the rest is just user error or forgetting to take a backup before testing something.
Proxmox, hello my old friend
Both of us have Proxmox clusters
Hm. So⦠the options are either
Whack-a-mole and ubiquiti level equipmentβ¦
Or vpn tunnels
Definitely pros and cons to both.
Ubiquiti you say.........
I used to use VMware with dreams on of a VCaP, then the recent crap happened and now VMWare can die in a fire.
I got the Ubiquiti equipment after seeing MANY suspicious connections in my old router logs being let through to my PC. They coincided with my security suite blocking exploit attempts. It now gets blocked at the router. I scored $1,400 worth of gear on eBay for $700.
Ebay is good for Ubi stuff, unless it breaks
Even with a warrenty, Unifi won't touch it if you're not the original purchaser
Yeah, I'm not terribly worried about that. It's solid gear.
I thought so too until a U6LR went bang on my wall
BANG?!
Yep, after spending a few hours and a lot of percussive persuasion to get it open afterwards, there was a lot of charing inside
I got a UDM-Pro, a 24 port PoE switch, and two U6 Pros for $700. The switch alone costs that much.
Any bugs or sign of why it shorted?
Nope, just a lot of SMCs that were no longer attached to the board... or in 1 piece
Tried to RMA it but since I brought it on Ebay and the seller has since vanished, I was out of luck. Replaced it with a U6 Pro directly from Unifi and I've had 0 issues
Ooof. That thing may have been having issues when they sold it.
UDM Pro, 24 PoE 250W, USW Flex, USW Flex Mini, 2x AC Lite, U6 Pro and 3x G5 Bullets
It was still sealed in box
Mine was "used", but except for a small scratch on the top of the UDM case, it's mint. Screen protectors are still on it, and SFP covers were on, too.
Oh my UDM was heavily used when I brought it, not had a single issue with it though
Not the best picture of all of it, but this is my setup
Everything was mint except for that scratch. Unused hardware pack for the APs, no labels or residue on them. Only issue was the wrong ears were shipped for one of them, but the seller sent me replacements free of charge.
Before I replaced the ears
I really need to get a patch panel for mine, but I'd want to run cables properly first and that's kinda hard in a rental
I have a strip on the bottom with power on the back, and a Raspberry Pi in a full heatsink case running PiVPN.
PiVPN? But you have a UDM that can VPN!
I waffled on it for a couple years. I regret not getting one sooner. Should be able to pick one up on Amazon for under 50 quid.
Yeah, lol. I'm getting that set up so I can allocate the Pi to something else.
I started with a Pi, then a laptop and now I have a mini server farm. Not pictured is another 4-5 rack servers sat in my office with another one potentially on it's way soon
My energy provider loves me
I have a 42U rack in my cellar I need to retire. The kit is too old to be useful now.
I plan to get a 42U when I've eventually got my own place, I'd also have no issues filling it aside from the power bill
I built a Proxmox box (fun to say aloud) with 128B of RAM and a decent proc, plus 9 spinning drives, two NVMe, and 2 SSD. I don't need the rack anymore. Lol. Just my network rack and that box. Everything else is virtual.
My main node in that picture has 384GB of RAM, 48 cores, 20x 1TB HDD, 4x 128GB SSD and 2x 256GB SSD
I do find most of my RAM vanishes to ZFS cache though
Sorry, didn't see your reply until now - but no, this is on 1.0 - I figured I'd wait for 1.1 . What I've done (which works - but a pain) - is run the game single player - then upload it to my server and let it run overnight.... then redownload the game single player... That way, no glitches - and the game technically runs 24/7
but no, reboots haven't fixed it π¦
8x 8TB HDD, 1x 4TB HDD, 2x 1TB SSD (RAID 1), and 2x NVMe that I can't remember the capacity of off the top of my head. I have a 6 core/12 thread AMD Ryzen at the heart of it. Lol. I need RAM and disk space more than I need cores or power.
64TB of the storage is so I can use Unraid to create a NAS of sorts. That's why I ended up with Proxmox. I've had limited success, and the whole needs some reliability testing. I had the original 1TB drive that held all my Proxmox info crash. Replaced it with a 4TB IronWolf.
my satisfactory server (running on ubuntu) keeps crashing
log file says:
[2025.05.31-02.56.54:889][ 0]LogCore: === Critical error: ===
Unhandled Exception: SIGSEGV: unaligned memory access (SIMD vectors?)
Unhandled Exception: SIGSEGV: unaligned memory access (SIMD vectors?)
^M
[2025.05.31-02.56.54:889][ 0]LogCore: Fatal error!
it will work for about 30-40 minutes, then crash
Unraid user too here π π
I'm trying to get it to work reliably to the extent that I would be willing to trust it.
Once I got mine up and running haven't had any issues, might be worth seeing if someone can help in their disc, heard about it from a old manager of mine who I was playing Satis with hahaha
Hey is there a fix for joining a friends dedicated server on 1.1
I try to load in and it crashes while loading everytime i try to join
Crash Report:
`Version: 410840, IsEditor: No, IsPerforceBuild: No, BuildConfiguration: Shipping, Launcher: Steam, NetMode: Client, IsUsingMods: No, IsSaveGameEdited: No
Unhandled Exception: EXCEPTION_ACCESS_VIOLATION reading address 0x00000000000000c0
FactoryGameSteam_ReliableMessagingTCP_Win64_Shipping!FReliableMessageProtocolTCPConnection::Tick() [D:\BuildAgent\work\SM_BT\UE4\FactoryGame\Plugins\ReliableMessaging\Source\ReliableMessagingTCP\Private\ReliableMessageProtocolTCPServer.cpp:263]
FactoryGameSteam_ReliableMessaging_Win64_Shipping!UReliableMessagingPlayerComponent::TickConnection() [D:\BuildAgent\work\SM_BT\UE4\FactoryGame\Plugins\ReliableMessaging\Source\ReliableMessaging\Private\ReliableMessagingPlayerComponent.cpp:67]
FactoryGameSteam_ReliableMessaging_Win64_Shipping!UReliableMessagingSubsystem::Tick() [D:\BuildAgent\work\SM_BT\UE4\FactoryGame\Plugins\ReliableMessaging\Source\ReliableMessaging\Private\ReliableMessagingSubsystem.cpp:89]
FactoryGameSteam_Engine_Win64_Shipping!FTickableGameObject::TickObjects() [D:\BuildAgent\work\SM_BT\UE4\Engine\Source\Runtime\Engine\Private\Tickable.cpp:153]
FactoryGameSteam_Engine_Win64_Shipping!UGameEngine::Tick() [D:\BuildAgent\work\SM_BT\UE4\Engine\Source\Runtime\Engine\Private\GameEngine.cpp:1864]
FactoryGameSteam_FactoryGame_Win64_Shipping!UFGGameEngine::Tick() [D:\BuildAgent\work\SM_BT\UE4\FactoryGame\Source\FactoryGame\Private\FGGameEngine.cpp:15]
FactoryGameSteam_Win64_Shipping!FEngineLoop::Tick() [D:\BuildAgent\work\SM_BT\UE4\Engine\Source\Runtime\Launch\Private\LaunchEngineLoop.cpp:5864]
FactoryGameSteam_Win64_Shipping!GuardedMain() [D:\BuildAgent\work\SM_BT\UE4\Engine\Source\Runtime\Launch\Private\Launch.cpp:188]
FactoryGameSteam_Win64_Shipping!GuardedMainWrapper() [D:\BuildAgent\work\SM_BT\UE4\Engine\Source\Runtime\Launch\Private\Windows\LaunchWindows.cpp:118]
FactoryGameSteam_Win64_Shipping!LaunchWindowsStartup() [D:\BuildAgent\work\SM_BT\UE4\Engine\Source\Runtime\Launch\Private\Windows\LaunchWindows.cpp:258]
FactoryGameSteam_Win64_Shipping!WinMain() [D:\BuildAgent\work\SM_BT\UE4\Engine\Source\Runtime\Launch\Private\Windows\LaunchWindows.cpp:298]
FactoryGameSteam_Win64_Shipping!__scrt_common_main_seh() [D:\a_work\1\s\src\vctools\crt\vcstartup\src\startup\exe_common.inl:288]
kernel32
ntdll`
anybody have a server i could join?
Is this the appropriate channel for asking to assistance with dedicated server setup? Run into a snag I can't quite untangle.
Probably
This is a reasonable place to ask π
I created a thread as I found similar content there, so rather than duplicate, #1378442930502631536
Is the satisfactory server more stable under Linux or Windows
and what would you reccomend for a stable server.. Mine keeps crashing after 30-40 minutes of my playing on there, and I am getting tired of it (I have to reboot it)... I was thinking of trying a windows VM
I've not tried it on Windows. 1.0 has been stable on Linux in my experience. 1.1 has had more issues.
I am on 1.0
Can you tell me about your linux VM or machine? I have mine set up for 16GB of RAM and 4 virtual processors
What's the failure mode? Any indications in the logs (for either the application or kernel?)
yeah its up above here...
Unhandled Exception: SIGSEGV: unaligned memory access (SIMD vectors?)
I also have crash report folders
Right, that's a segmentation fault β tried to access memory it didn't have permission to.
I did a ubuntu (minimal) install 24.0.2 LTS
That sounds reasonable; I'm using Debian which is not dissimilar.
I believe I have the latest 1.0 release binaries (I am using steamcmd). . If you tell me the command to run I can report back the version, etc...
I figured it would be best to run it under Linux, but I think I may try a windows server VM instead (btw this is running as a VM in proxmox)
The server machine has like 64GB of RAM and like a 8 core processor 3.x ghz ,etc..
(the physical server that is)
Raid5 adaptec controller, etc..
Researching, might be related to OpenSSL libraries being in conflict, per https://github.com/getsentry/sentry-unreal/issues/259#issuecomment-1553372222
I also wonder whether the log suggests that this error happens when it's trying to report the crash? It could be the actual triggering event was something else.
I had my systemd service set to Restart=on-failure
but what would happen is once the crash happened
it would restart and restart and restart and restart over and over again
so I removed the Restart in the systemd service definition, and now I just reboot the VM when it crashes, and try again
... Im not sure what to do about the OpenSSL libraries ...
You may wish to tweak the .service file to add a restart delay.
had RestartSec=10
... is that too short ?
Im kinda doubtful is this ssl thing, that article was written back in 2023
those types of security libraries are updated all the time
That can be the problem. If a game was statically compiled with OpenSSL 1.1.x, but a seperate library tries to link in OpenSSL 3.x (supplied in Ubuntu 24.04), they can trip over each other.
okay
But you're right to suspect a red herring.
.. Is there a solution in there (in that article)?
It doesn't appear to be a problem unique to you:
https://questions.satisfactorygame.com/post/66e884d7772a987f4a8b3f4c
That article talks about a particular add-on for Unreal engine programmers, and how they avoided a version clash problem. The solution doesn't apply here.
(Or, if it does, we may need Coffee Stain themselves to figure out the fix.)
what is the command to report the satisfactory server version with steam command ?
(btw I am not using docker or anything, just a VM with this running, and a minimal install at that)..
What kind of VM? Proxmox or similar?
proxmox ya
I note there is this point on the wiki:
On the Proxmox hypervisor specifically, check that the CPU type defined for your VM is not KVM64. This processor type will cause the Satisfactory Dedicated Server to crash on load. Instead of the KVM processor type, select host, or another processor type to resolve the crashing. This may also affect other hypervisors depending on your configuration, if you have the processor type KVM set.
(I would link to it, but the moderation auto-bot doesn't like it.)
I bet thats it
I don't know if that's a current problem, but it's worth looking at.
(Several of the comments on the wiki go as far back as Update 5)
But yes, if it's an architecture emulation subtlety, then this would be a nice and, dare I say it, satisfying fix.
in proxmox it shows it as x86-64-v2-AES
Mmm, that's kinda modern.
that was for "processors", for "machine" it is showing "Default (i440fx)
Now there is not much of a graphics card in that server there, but I dont think the server requires anything for a GPU does it?
Per the Proxmox wiki, the levels are:
kvm64 (x86-64-v1): Compatible with Intel CPU >= Pentium 4, AMD CPU >= Phenom.
x86-64-v2: Compatible with Intel CPU >= Nehalem, AMD CPU >= Opteron_G3. Added CPU flags compared to x86-64-v1: +cx16, +lahf-lm, +popcnt, +pni, +sse4.1, +sse4.2, +ssse3.
x86-64-v2-AES: Compatible with Intel CPU >= Westmere, AMD CPU >= Opteron_G4. Added CPU flags compared to x86-64-v2: +aes.
x86-64-v3: Compatible with Intel CPU >= Broadwell, AMD CPU >= EPYC. Added CPU flags compared to x86-64-v2-AES: +avx, +avx2, +bmi1, +bmi2, +f16c, +fma, +movbe, +xsave.
x86-64-v4: Compatible with Intel CPU >= Skylake, AMD CPU >= EPYC v4 Genoa. Added CPU flags compared to x86-64-v3: +avx512f, +avx512bw, +avx512cd, +avx512dq, +avx512vl.
well, the physical CPU on that server is: 8 x Intel(R) Core(TM) i7-3820 CPU @ 3.60GHz (1 Socket
i440 is the older-style chipset that Proxmox emulates; there's Q35 as a more modern alternative, but it's unlikely to be too significant.
You don't need a (virtual) GPU.
Is there any chance I can submit my crash files to someone at Coffeestain for analysis ?
It should autosubmit if it can.
Should I give it more cores ?
That won't help with crashing.
I'm trying to determine whether your CPU can run at x86-64-v3 level. That adds the AVX and AVX2 instruction sets, which are a type of SIMD instruction...
Perhaps what I should do is choose "Host" for the CPU type, rather than let proxmox sleect it as V2-AES like it did for this VM
I am not all that great with proxmox, but maybe choosing "Host" will make the VM run with the most compatible CPU (I imagine it will try to match the Physical CPU ?)
Host will make all of the CPU features of the VM host available, it's a good option if you're not going to need to hot-migrate the VM to another machine with a lesser capability CPU.
or choose the "V4" option (intead of V2)
So, what Proxmox is doing is running up a virtual machine. It doesn't automatically pass through all the details of the host machine's CPU, and limits it to the -v2 compatibility standard.
more modern CPU ?
-v4 is more modern, but I'm not certain your CPU is actually new enough to provide all the features that compatibility level implies.
I see
This is only really important if the VM is going to be migrating from one Proxmox VM host to another without being rebooted first.
If you have a single-server Proxmox machine rather than a bigger cluster, pick host.
(If you have a cluster, but all the machines are the same class, host is also a reasonable option then.)
okay lets try that (can I just switch this VM, or do I need to re-install a new VM and start over for this?).. I am not sure how Ubuntu is going to feel about a new CPU inside the VM
probably need to re-compile the kernel or something
No reinstall or other changes required, only a reboot.
(Linux is generally very good at Just Copingβ’ with whatever hardware it finds to hand.)
You may need to turn the VM off before Proxmos will let you change the setting on that VM.
This is a per-VM setting.
Boot it up, see what happens. π
(It could still crash, but at least we'll have eliminated one possibility. π )
(it literally boots in like 2-3 seconds)
Yeah, nice, isn't it? π
(however satisfactory service takes a bit to start up)
well this machine (physical server) is a bit of a beast
I sorta stole it from work π ... they were gonna throw it out ...
Rate-limiting step is usually storage subsystem. If SSD, usually goes zoom
But it sucks down 130 watts so.... probably $25 a month just in electric for it...
It's not stealing, it's recycling. π
Yeah, I've recently rescued some old VM hosts from a friends' work for the cost of shipping. Nervous about the power draw for 6U of chonk
... but I might be able to get solar panels and a battery soon, which would help a lot ...
6U damn...
... Wow... the satisfactory service failed to even load this time ...
came up with ualigned memory access right away
[2025.05.31-21.01.37:699][ 85]LogCore: === Critical error: ===
Unhandled Exception: SIGSEGV: unaligned memory access (SIMD vectors?)
^M
[2025.05.31-21.01.37:699][ 85]LogCore: Fatal error!
(booo)
I wonder if I should just use a different linux distro
(I am most used to ubuntu)
Or I wonder if I shouldn't use minimal, and actually let it install a GUI
GUI is unlikely to help
(Might help you, but the game definitely doesn't require it.)
Do me a favour: run dpkg-query -l | grep curl and tell me the output
Or (again) just install under windows server (I could do that I suppose)
(This will list all of the distribution packages installed on your machine that have curl in the name.)
root@satisfactory2:/home/xxx# dpkg-query -l | grep curl
ii curl 8.5.0-2ubuntu10.6 amd64 command line tool for transferring data with URL syntax
ii libcurl3t64-gnutls:amd64 8.5.0-2ubuntu10.6 amd64 easy-to-use client-side URL transfer library (GnuTLS flav our)
ii libcurl4t64:amd64 8.5.0-2ubuntu10.6 amd64 easy-to-use client-side URL transfer library (OpenSSL fla vour)
Okay, so two different cURL versions, 3.x is GnuTLS, 4.x is OpenSSL β mine's the same.
(Though you've got the t64 variants which use a different time_t, don't think that's likely to trip you up..)
(and I sorta doxxed myself there, oh well)
Its really too bad, this is a nice server to run satisfactory on
I had it on a way underpowered machien before but
it didnt crash like this
It's an older CPU, but I wouldn't expect it to crash hard. :/
The other machine was an intel NUC with a n100 processor
so newer processor I suppose
but really meant for light loads..
... Also my save got more complex
oh great now I screwed myself... I set the CPU to the v4 one and tried to boot, and it failed.. so its stopped now, but now I cannot edit the CPU anymore (As if the machine were on)
Try forcing a VM power-off.
(host will provide maximum capability, I'd leave it on that setting.)
okay I think its back
well with host it failed... so Im gonna switch it back to what I had and see if at least the server starts up
You might like to try adding a line to your satisfactory.service unit file with something like this, to ensure that the game files are validated on start-up:
ExecStartPre=/home/steam/Steam/steamcmd.sh +force_install_dir /home/steam/opt/satisfactory-dedicated-server +login anonymous +app_update 1690800 -beta public -validate +quit
(Adjust paths to match your system as necessary.)
yeah I have something likke that
ExecStartPre=/usr/games/steamcmd +@sSteamCmdForcePlatformType linux +force_install_dir /home/satisfactory/server +login anonymous +app_update 1690800 +quit
... although I see... pehraps I should add -validate ?
just before +quit?
-validate +quit
right, so I put that in there and now its taking its time starting up (I take that as a good thing)
I think it took too long... the systemd console log shows
May 31 16:25:59 satisfactory2 systemd[1]: satisfactoryserver.service: start-pre operation timed out. Terminating.
so I am just gonna run that PRE command from the command line directly?
anyhow thank you for your help
I dont think Im gonna get this solved today, at least not under this installation
nope havent been able to null it out... didnt happen on 1.0 but its every 30 min ish on 1.1
anyone have a server open
those that do list them in #1201555265942724758
I hate to say it, but good luck with that - i wouldn't let anyone I didn't know on my server... opens up for trolling (since everything is serverwide - not individual....) - best bet, find friends....
you cant do v4
really if you are not going to migrate, host will make it way easier
the other option may be manual tunning capabilities
]LogNet: AddClientConnection: Added client connection: [UNetConnection] RemoteAddr: 87.225.97.100:57591, Name: IpConnection_2147434921, Driver: GameNetDriver FGDSIpNetDriver_2147460400, IsServer: YES, PC: NULL, Owner: NULL, UniqueId: INVALID
how to fix?
I added to my unit file:
TimeoutStartSec=1hr
My dedicated server doesn't show as online in the server manager. I'm on experimental and have ports 7777/8888 open for TCP and UDP in windows firewall. The server connects when I use localhost, so I assume it's just a network issue, but I figured I'll ask as I had it running perfectly fine before
auto-connect mode only works a few times in a row and then it doesn't place... restarting the server fixes this (allowing us to place a few more auto-connected blueprints before happening again). Anyone else seeing this issue?
nevermind, I was just using the wrong IP
it seems to have changed from last time I ran my dedicated server
having trouble with a server, buddy is creating it on his pc but he cant join and neither can i
any tips?
getting server appears to be offline notice
make sure you are using current port assignments. It's changed a few times over its lifetime
!wikisearch Dedicated_server
This page provides basic instructions on setting up and operating a Satisfactory dedicated server.
Dedicated server binaries are available for 64bit Windows and Linux systems. It can be installed using the Steam client (where it is listed in the user's library as a "tool"), its command line sibling SteamCMD...
even as we speak, there is a diff in port assignments between stable public and EXP branches
trying to just do a session, but cant join with one on steam and one on epic
that isnt a server
i know that, we stopped trying the server
@dwm TimeoutStartSec to one hour?
anyhow for those that were following me earlier... I did get the service to start up with my CPU type set to Host in ProxMox...
... I think its more stable now... I played it last night and today and it hasn't crashed on me yet... It did crash last night (without anyone on it) at 3am, but I think that is when the server restarts itself and I guess it wasnt able to start upa again? (I changed it from 12am to 3am)
So I am somewhat confident that it is working better now, although if it crashes again I will come back in here and let you know... So I thank everyone here for all the help ..
When the server shuts down each night (defaults to midnight) what does it do? Does it know how to start itself up again or do I need a restart in the systemd.service file to tell it to restart on crash ?
Anyhow, it still may be too soon to know if it is really stable or not... And of course I will take the update to 1.1 on its release on June 10 ...
(I assume the update will apply to the server too)
You could just raise the autosave count and set up periodic backups of the save files themselves. Random people can be fun.
heh - fair enough π
Anybody able to help with getting certs to work with self hosted servers? I have a cert for satisfactory.my.site loaded. I can see in the satisfactory load the cert on server startup. The server remains unreachable at satisfactory.my.site. In Client it's just stuck at Server name pending Not Authenticated This server appears to be offiline. The server log shows a spam of NotifyAcceptingConnection from ipconnecting.
Hello... So I believe that the "Host" setting in proxmox for the CPU solved most of my stability problems with the server... However now I have another wierd issue. The server crashes when it attempts to start back up again after its nightly restart process (the one that is built into the game when it restarts itself every 24 hours)..
I think its a legitimate crash because I see this in the error logs:
[2025.06.02-08.00.15:795][166]LogCore: Fatal error: [File:.\Runtime/Core/Private/HAL/MallocBinned2.cpp] [Line: 1141]
FMallocBinned2 Attempt to free an unrecognized block 0x10000000000000
(the 8am is showing there because that is 3am central time when I set the option in the server manager to restart itself)
Has anyone seen this one before?
.. Looks like its coming from the unreal engine
Do you have the server setup as a service in systemd?
It is, and I have it set up to not restart itself on crash at the moment (Restart=on-failure is commented out), however the server should be able to restart itself after its nightly shutdown cycle, yes?
... another workaround would be just to reboot the server after 3am each night I suppose, it seems to start itself up okay because of the systemd
anyhow, if I did set up Restart=on-failure, what should I put for RestartSec (Restart seconds ?)
For sure, on-failure will only attempt to restart the server if the exit is non-zero, i.e. anything else but zero once the process dies, systemd will flag and then attempt to restart the service.
I recommend leaving it at default with 60, can you send me a screenshot or a copy-paste of your .service file?
sure
I only say 60 because if it hangs after crash give the kernel time to dispose of it, can experiment with it though
can I use pastbin or something?
yeah or paste it directly here up to you
and are you able to get your entire log as well?
yeah.. I have the game log, and I have the corresponding "crashes" foldedr
(seems that the crashes folder has a copy of the log from the time it crashed)
I can TAR that up
Is there a reason you have +@sSteamCmdForcePlatformType included in your execstart
not that I can think of (I copied the .service from a guide for the most part, so I don't know exactly what all of it is for)
Which guide did you use? Official wiki has a different service file
https://satisfactory.wiki.gg/wiki/Dedicated_servers/Running_as_a_Service
I too, am running proxmox with host as cpu type on EPYC CPU for the host, and i have everything setup as per official wiki guide and don't have this issue so either its configuration issue or environment issue i'd say
Me too, that's why i think you might just have a small config error in your systemd unit file
i dont know why forcing the platform to linux makes sense here
,,, If there is a better guide I should be following Id like to do that ...
https://satisfactory.wiki.gg/wiki/Dedicated_servers
Follow the official guide ^
Environment=LD_LIBRARY_PATH=/home/satisfactory/server:$LD_LIBRARY_PATH
This is also pointless, not sure which guide you used but i would start again on official routes
yeah I think I may have seen that before, but that guide doesn't really have an example of a full systemd.service file, which is probably why I shopped it from somewhere else ...
i would do a fresh install if it's possible just in case
but ytou should be fine just altering your systemd file and tidying everything
yeah
.... I was having all sorts of crashing issues before, but that was before I set up proxmox cpu type as host (it didnt work the first time I tried it, I don't know why, but it is working now)
I have a 12 year old CPU
but its 8core and a a fast i7 processir
by default proxmox will use qemu type cpu's so likely you'll be using a cpu type that will print kvm64 on lscpu or similar tools
official guide states kvm64 cpu won't work which makes sense because the game will need to use specific instruction sets which kvm64 type cpus dont have
8 x Intel(R) Core(TM) i7-3820 CPU @ 3.60GHz (1 Socket)
Thus changing it to host cpu on the vm will make those instructions available to the vm and the game can do what it needs
... that cpu should be good enough ...
it is, its fine
and the server has 64GB of ram, so I dedicated 16GB to satisfactory server
But your right, I should be using the official guide, as a lot of stuff out there is way outdated (or just plain wrong)...
It's fine, it can happen, let's just hope it works or gets you one step closer to a solution π
... although I don't see the port number defined in the systemd service there
Read from the top of the dedicated server section of the wiki
It will tell you the answers, I.e default port is 7777 if you donβt specify any
If you want to specify your own head down to the command line arguments section to see what you can specify with the script
15000 and 15777 are no longer requied
(i think they used to be, which is why they show up in other guides still)
Oh yeah I just realised you have vanilla settings in your old config
ah, and the guide does say that
Lightbulb moment aye
"As of Patch 1.0, port numbers 15000 and 15777 are no longer used. At all. Sometimes port 8888 will need to be opened."
what is 8888 ?
not a problem
Default ports:
1.0: 7777 TCP & UDP
1.1: 7777 TCP & UDP, 8888 TCP
ah, so we need 8888 tcp for 1.1
I guess I am having a hard time understanding from the 1.1.0.0 patch notes, do we need 1 or 2 ports (i.e. 7777 and 8888, or just 8888) ? I think this needs some clarification ...
from https://satisfactory.wiki.gg/wiki/Patch_1.1.0.0 -> "If hosting a single server, port 8888 TCP must be open by default.", it doesn't mention 7777 at all
7777 TCP nad UDP and 8888 as TCP, for Experi
Because there is no changes to 7777, so I think you can safely assume it's business as usual unless they specify otherwise.
Are there any go developers here btw?
... I would hope that the devs check in on this forum channel from time to time ...
They probably just look at the site ones
I can see that the community managers are online (Sometimes) but I certainly don't want to be bugging them ...
If running Experi you need these
okay, so once released, I Shouldn't need 8888 then, and can stick with the default 7777 port ?
You did not read it did you, you need both
.. I must be missing it
I misread that one sorry lol
okay
Yeah once out of Experi you remove 8888
https://github.com/Xeckt/gofactory
Looking and accepting contributors for the new Go project if anyone is interested. Will have most of the API integrated in the next couple days then will add unit tests and move onto making the CLI.
API library and CLI are bundled individually, so API itself is a client lib and CLI will use the client lib to do the work.
That sounds interesting.... What would the client do?
(sadly, I know nothing about Go, I am a C#/VS/SQL dev thou)
Readme of the project will give you a quick overview
not interested on making the readme perfect rn though
Can you use the API to bring up the crash logs?
(cause that would be convienient, lol )
... and perhaps move blueprints in and out of the server? (that would be convienient too)
don't think the api has that function but the cli tool i make certainly could output logs
I can do all those things now, its just a PITA, would be nice to have a tool do that...
i can look into that also, the cli tool is intended to give as much management over the server as possible
how would you want that to behave?
Well, pick a save, see the blueprints there, copy them in and out ...
Or copy blueprints from save to save ? , etc...
Where is the best guide for the API ?
im using the openapi spec atm https://github.com/satisfactory-oas/spec
I suppoe in theory even if the API Doesn't support it, you could probably do something like that using SCP or something (assuming a linux server) ..
excellent, this is awesome... Had no idea until now this existed
... Doesn't seem (at first glance) that the API supports blueprints though (which is a huge miss IMO)... I can see you can D/L the save game thou...
Satisfactory sorts keeps the blueprint files seperate from the save game (but still has them set up on a per-save basis) ...
I mean, its not a big deal to go into the server and copy them around, its just too bad they have the ability to upload and download the save file, but not the blueprints, which you can d/l from other players, etc...
its more lilkely that folks are gonna share blueprints than they are to share their world files...
... and for folks that are unable to get around a linux command prompt and use SCP, etc... It would be nice for them to retrieve their blueprints so they can use them in other saves, or share them, etc...
Well it is only v1 api so i'm sure this will expand a tonne later, and that will become part of it.
Would be cool because with any api library you could make two concurrent connections between two different servers and then add a bunch of your own logic to copy blueprints to one another.
yep
in the meantime though i can look into copying and compressing blueprints to a zip in the cli tool if it's meaningful
most folks can handle downloads and uploads from their PC, but few folks can get around in unix ..
i mean if someone is hosting their server on a linux dedi/vm i would presume they might be able to, but assumption can be a killer π
well yes, they should be able to, but the folks that are playing on their server and have blueprints to share, that is the bottleneck...
Most folks cannot even locate their blueprint files on their PC ...
Yeah i mean for now I will look into making my cli tool copy out your blueprints and then zip them up, but ill need to dig a bit deeper on the nuances as im pretty sure it's session based so need to find a reliable way to identify sessions
only thing is my tool is for those comfortable with the CLI, it's not a GUI tool which is where you C# WPF folks can come in, I'm sure there is some out there π
right
and if the API could get to the log files, that would be nice too (I don't see that in the API spec either ) ...
this is why i've bundled mine into two seperate projects
Was the high cpu load on ded. Servers fixed in the new Patch today?
Any dedicated server open avaiable? Is there any list of servers?
Sorry to keep posting this, hoping someone at some point will notice and know what I can do to troubleshoot this. I can not access my server through my VPS reverse proxy. The WG tunnel ip hits the server with NotifyAcceptingConnection, but clients show the server offline.
I just started back playing this game, Anyone else having an issue with the game crashing when it auto saves?
How are you trying to connect to the game WG or Reverse Proxy?
High Server CPU Load looks like its fixed in the Dedicated Server (Experimental) Update. THX! Was a pain to run 12 cores at 100%
User ---> VPS running Pangolin ---> Newt Tunnel ---> Server
This works with every other game I have tried, but Satisfactory is being a real pill.
Thx to the devs π
On the User end it's stuck at the Server Name Pending. Initially I thought it was an ssl cert issue so I loaded the certs from Pangolin into the Satisfactory server, but the behavior remained unchanged. The log is sparse, shows no errors, just spam of the NotifyAcceptingConnection to the WG ip whenever someone tries to connect.
I have not messed with pangolin very much. I installed it when i found it could tunnel TCP/UDP unlike CF tunnels. I guess i may not be the best person to ask with that being said.
It is very odd that other games work. has crowdsec made a decision or authentication getting in the way?
I don't have crowdsec enabled at the moment as I wanted to remove potential issues. So I know it's not that. And the server works just fine on its local ip so I know it's not the server side issue.
Does your firewall see the connections from the VPS are they getting blocked?
Not sure what you mean by that. Pangolin has port 7777 open for both tcp and udp.
Are you comfortable giving me the address to the server so i can try to connect from it?
LogNet: NotifyAcceptingConnection accepted from: 172.18.0.2:59865
LogNet: NotifyAcceptingConnection accepted from: 172.18.0.2:56003
19:26:09
LogNet: NotifyAcceptingConnection accepted from: 172.18.0.2:39744
LogNet: NotifyAcceptingConnection accepted from: 172.18.0.2:50026
19:26:19
LogNet: NotifyAcceptingConnection accepted from: 172.18.0.2:34290
LogNet: NotifyAcceptingConnection accepted from: 172.18.0.2:35440```
That's you.
Hey does anyone know what this process is? Satisfactory is the only dedicated steam server downloaded on this machine leading me to think either steamCMD or the dedicated server download is compromised.
No server is running and this process is requesting 1200% cpu
reboot?
are you using docker?
No
wtf
Downloaded steamCMD from repository
Used the wiki page to install the dedicated server
What port is WG or pangolin using WG default is 51820. Is the 1.0 or expermental?
1.0 and lemme verify but I am pretty sure it's 59865. Need to look in the traefik yml.
i mean it seems like the cores are beeing reserved but yea strange usage....
Thatβs the issue itβs not just reserved itβs using the processing power my tower sounds like a jet engine lmfao
Nope this is native
It could be something in SteamCMD cause Iβve purged the whole user to get it to stop but after making a new one it comes back
it can be a stuck process of install or update...
can you kill the process and see what happens?
You're running traefik and pangolin?
It comes back on in ~5 mins
Pangolin is built on top of traefik
okay i was gonna say.. lol
Pangolin is more or less a fancy UI for Gerbil and Traefik.
Killed & timestamped
This is not experimental right?
Negative. It's 1.0
Itβs already back π
2 minutes later
I can see the connection being made to your IP address
even uninstalling steam cmd?
Yup, it's getting there and then ?????? Fun one isn't it, lol.
Iβll nuke it and see
the last option is to clean install the OS
this is why i like VMs
I can see the connections on my end too.
My side is showing 51820 is rejecting connections
I should hope so. That's only open from the VPS to me. Not to the outside world.
Iβd really not like to do that plus Iβd have to wait til I get home and have physical access to the machine
True but why not use two computers when I got them π
Good most people just open them wide lmao
yea thats true, long propcess
Also a lot of other data lost
Ask me how I learned not to do that.
back it up
If itβs malware it wouldnβt matter
I feel like the other chat thats happening might be finding out.
thats true but coming from steam.....
π₯Ά
I was going say Malware that mascarades as Steam is rather common.
anyways going to sleep ill see messages tomorrow if anything changed, its 2 am lol
Thanks for your help gang
Iβll try and talk to some people who are familiar with SteamCMD
yea
At this point i might reconsider nuking the satisfactory instance as it seems to be the only thing not playing nicely. i can see the connection complete even via 7777
I thought maybe UDP issue with teh VPS, but Factorio works fine and that requires a UDP connection.
i mainly use pterodactyl, but that has its bugs aswell
Sad that project ground to a halt. AMP is worth the $10 imo. It's what I use.
I've tried nuking and recreating it a few times alas the same result.
nah bro i have all working, also homelabing is about trying to solve issues to learn
still pelikan is on development
Agreed. Hence my presence here right now, lol. I really want to get this server upa nd running but it's driving me batty.
hahaha
i gave up on VPNS for hosting servers
some people need acces to the pannel aswell so vpn is a no go
I was doing it the no so safe way for a while until I found Pangolin. It's a little rough around the edges to get set up, but it works. And it's free
i had enough headaches making pterodactyl work with a reverse proxy and finlly got it working with a CF tunnel
Thanks for taking a look, if you can think of anything else to try I would appreciate a ping.
well i should go to sleep, nice to talk to you
Take it easy
Purged and it came back π
that is quacking like a malware duck my friend.
Purging the user gets rid of it however
Imma try reinstalling steamCMD but not installing satisfactory to see if the process returns
Another weird thing is looking this up only yields 1 person having the same issue and also making a satisfactory server https://www.reddit.com/r/Steam/comments/1fjzysl/steamcmd_on_ubuntu_installed_and_steam_user_runs/
Anyone ever experience lag spikes on multiplayer when crouch jumping?
I just don't understand why I am having such a hard time re-installing (updating) my satisfactory server
I wanted to sort of start fresh, so Irenamed my server folder to something else, and now I am trying to re-install using this:
steamcmd +force_install_dir /home/satisfactory/server +login anonymous +app_update 1690800 validate +quit
... now I am switched to the (user) satisfactory (a user I created for this), and I am doing this in the satisfactory home directory
hence: /home/satisfactory
I am following the official guide in the wiki
when I run that command, one of three things happens
- I get this error : IPC function call IClientAppManager::GetUpdateInfo took too long: 56 msec
.... and it quits
or
well, It was doing a couple of different errors (seems l cannot find those other errors I copied down), but typically I get that error, and it stops the install
then I remove all files from /home/satisfactory/server , and start over
what is that error?
@nocturne crescent Found the answer and its a bummer. Can't add traefik labels to an AMP Instance yet that I can see. https://github.com/wolveix/satisfactory-server/discussions/241
Oh So its a docker network issue. I just rented a vps and was going to try to set it up lol
I tried that, but according to the guide above you need to add pretty specific labels and I can't find any AMP documentation on what that button actually does.
It's both a docker network issue and a certificate issue baked into a crap pie.
I bet you have to use the console to edit the docker compose file.
No dice. Just happened to catch an admin from AMP. traefik inplementaion isn't any good and he didn't recommend even trying to get it to work. I am probably just going to have to run this in its own lxc container outside AMP for the time being.
I run all mine in lxc or a debian vm. I have amp but that's more for messing around or being able to spin up a server for a game with like 2 clicks.
Yah, I like AMP so my friend group can create servers without me having to do everything. I give them access to Pangolin and my VPS and then really the only thing that can fail that will always require my intervention is the newt container. It works for everything else, just not Satisfactory, lol. Cest la vie. Nothing is bullet proof.
i just read the thread on the CC discord. Greelan is a knowledge sob. I was going to say i wonder if he could assist but.... yea..
Yah, the lack of info in the AMP console had me chasing down the wrong lead for a few days, but he helped me eliminate that possibility.
As he pointed out the container just needs to parse the whole log and not just parts of it, but that's a seperate issue.
I got it! Binding the Satisfactory instance in AMP to the host IP(rather than 0.0.0.0) and the tunnel to the host IP and I am cooking with gas.
Heck Yea, Nice job. I'm gonna cancel this addiotional VPS order lol. I was getting a completly seperate instance going. I guess i'll play with pangolin for the month i have the server lol
I have no idea why that works, for what it is worth.
dammm
let us know how it goes
Hi peeps, my api client library is now fully integrated with the latest v1 dedicated server api, working toward v1.0.0 and tidying things up. You can check it here, and welcome to start using it if it's up your alley.
The CLI tool for people who host will be in the works as soon as v1 is released
Hi yall Iβm trying to create a satisfactory server on Ubuntu machine but Iβm to the point of starting the service but it says it is masked. I tried unmasking but it deletes my systemd directory when doing so. Iβm following the guide posted by Astral internet
Little preview of how the cli tool will report back from the API. Will feature non-api, direct server related functions that will be useful for server hosters as well.
v1 of the API client library is released with full HTTPS API integration. Next step is to integrate a lightweight query api client library for v2.
All fronts will be supported for server management, except, well, mods.. for now.
Hello, I have been looking into the reports we have received regarding High CPU usage on dedicated servers after few hours of it running. I would like to know if anyone has ran into this issue where the server was hosted on windows OS ? Internally I have been seeing this issue on linux-dedicatedServer [When client disconnects and server is running for couple of hours].
https://questions.satisfactorygame.com/post/67efeb096b7c573196363779
https://questions.satisfactorygame.com/post/68243a4a6b7c57319637a60b
https://questions.satisfactorygame.com/post/67f4ff9d6b7c573196365303
This is isolated to 1.1 experimental it appears?
I can spin up a 1.1 instance to observe it's behavior vs my 1.0 instance running identical save files.
Running on a windows server, and I've crashed twice with experimental 05 - and my taskbar still keeps being blank randomly when i log in (I have to scroll the taskbar to another, and then scroll back for it to be restored....) - but other than that, 1.1 seems stable (I'm more happy the terrain bug where it wasn't updating the client has been fixed....) - haven't checked to see if the water has been fixed in the rocky desert (the cave near the sam specifically)
but, I haven't noticed anything like the links so far....
yep it is 1.1
Can anyone tell me what the problem is?LogHttpListener: Error: Error accepting expected connection [0] SE_NO_ERROR and LogReliableMessaging: Error: Chunk header refers to an invalid channel id. A message header wasn't sent before this chunk. Closing connection.
Not that anyone really cares here but in case someone might in the future. My api library now almost has full support for the lightweight query api and will be featured in the api library as a whole, so LWQ and HTTPS api bundled into a single library. Pretty cool learning experience for me π
I spun up a 1.1 container yesterday. I'm not yet experiencing this on docker containers. It's still running to see what happens as it gets more uptime. Currently at about 24 hours.
I am going to create a VM with the server on it next to see if I can't recreate the issue outside of docker.
Hey guys i keep getting connection to host lost, What could cause the problem? the server is in my home so its LAN same VLAN any no firewall rules to deny it, Im using pterodactyl running the game in docker container but the issue didnt happen on my last server it only happened when i port forwarded to the server i started getting these issues altho i tried connect via local ip and via public ip same issue
So you're getting kicked off of the server only after you forwarded the ports?
Yeah for some reasons but before i did it was fine
Remove or disable the port forward rule and see if the issue persists.
Aren't there any debugging/error logs that i can see? other than the usual logs for joined and disconnected?
I will try it after since my friends are currently on thanks π
Is it kicking them off too?
Not really
So just you are getting kicked locally when on the same vlan as the server?
Yeah same vlan and nothing in the firewall that will cause the issue and it is on the same L2 too
I take it your PC is hardwired not wifi?
Correct yeah, same L2 same switch with the server and the PC
Hmmmm.. is your PC using alot/most of its resources running the game?
Nope, networking is usually low and CPU/GPU doesnt go above 60% I will try to disable port forwarding and see, my friends arent getting disconnected maybe because they aren't playing as much as i do so there is not enough time to get DC'd but I will try and see, even tho other services arent getting me disconnected on the same ip i have multiple games servers running on it
Are you playing on 1.1 or 1.0
1.0
Hmmmm
Yeah that is exactly what i said, i had nothing in mind networking is in check latest update as well, everything is working except the random DC like its not even public network that i could say maybe the WAN
The game and server should work regardless of wan status though because it's local and on the same LAN
My Server (1.1 Experimental) runs in a LXC Container (Proxmox) with Debian 11. Attached 6 cores with 16GB Memory and 64GB Storage. Host is a i5-11400 with 96GB of Ram. Same Problem here, after e few hours - no one is connected/playing AND Server Stops the Game without players - High CPU Load.
I was playing from ~22.00 - 01.30 and again from ~08.00 - 10.00 - there is a spike around 14.00 - very Strange π
ive only ever used local server hosted for my friends to play on....but for 1.1 drop id like to have a dedicated one. im not sure my rasberry PI could handle it tho. would it be able to? or should i get a month or 2 of paid hosting site. how simple is that process if so?
If you go paid, I'd rent from a gamerserver company directly. Those make it really painless to setup. If you go VPS you'll have to configure it all yourself.
can relate, no player joined the server at all but cpu spiked several times:
Any good suggestions of reputable ones?
Has either of you guys been able to run the dedicated server through CasaOS? @pulsar cove @pine grove
Or anyone of you guys
I can connect it normally locally, but, remotelly it does not work at all.
As you can see, does not even get the name...
(I used another ISP connection to test the remote ip from my homelab)
From outside in run into the same Problem, even with portforwarding (Fritzbox). But i dont care about it and closed the ports. Connection via VPN works (wireguard) π
I mean, VPN would be a solution, but, not practical for my friends unfortunatelly.
True
Look at my cpu usage. 6 cores from a 11400 at 20% with one Player connected. I dont think a raspi 4/5 can handle that. π₯Ί
Did you change the ports, or are they standard? You may need to use host as the network type.
is there a dedicated server tool out there thats just an application of some variety? with a GUI and simple interface. An options menu with checkboxes for different toggles, for example, something simple and streamlined like that
I am having some issues with setting up a public server on my PC. I have downloaded the steam server and enabled port forwarding to my pc internal IP in the deco app for my tp-link mesh system
internal IP server works fine ingame, but whenever I assign my public IPadress when trying to create server ingame, I just get a message saying it is offline and servername pending and cant authenticate
you router might not allow connections to your external IP from an internal. best test it with a system outside your network
use phone wifi sharing and try through that?
if you use mobile data for upllink
@whole meadow please let us know what ports are needed to forwarding. ATM i forward port 7777 udp/tcp and 8888 tcp and cannot connect to my server without vpn
Hi, i created a server and tried to increase the max player cap to play with friends (we are 5) but when i did the whole change game.ini file and type how many players, it didnt change it. Later i tried to do a local one (did a new world, no server) it did rise the player cap to 5
anyone know how to fix it for the dedicated server?
Dont think I understood properly. How would I go about doing that?
ask a friend to connect, use an online port-scanner to check or, as you said, use phone hotspot (but make sure to use mobile data, not wifi, as uplink)
If I use an online scanner with my public Ip adress I just get a timeout on all portsπ€
I use the one I get on whatsmyip and it matches the IP in the routers app
Or am I an utter moron ingame when I try to create a server with my public IP instead of my local pc
And when my friend tries to log on he should get my public?
well you do need to test your public IP if you friend wants to join
for yourself your private/internal is fine
any idea why this isnt working? Trying to run nssm.exe install SatisfactoryServerService in the directory containing nssm.exe, as per the wiki instructions
try .\nssm.exe
ok, that seemed to work. Any idea why that seemed to be the case, and does the wiki need updating to reflect that?
So if I setup my server with my internal ip it should work for him if he uses my public?
Powershell seems to look into PATH by default, with the .\ you tell it to look into the local folder.
What dou you mean when you say "setup"?
Ah well. Just in game on manage servers with the steam clients own server tool
you should bind it to your internal IP and forward to that in your router, you can't bind it to your public IP on your PC
Ye think that fact occured to me a bit late and that I misunderstood some stuff
Yeah, if you mean with what you add it to the server manager than yes. You use the internal IP and everybody else needs to use your public IP (without a VPN setup)
so I apparently need to run powershell as admin within the folder of nssm. Im having trouble figuring out how to do that, any tips?
Launch powershell as admin, then navigate to the folder
Right click on the start menu icon > launch powershell/terminal as administrator
ok, after looking up how to navigate through powershell (cd), i was able to get progress. except now im running into this
SatisfactoryServerService: START: The system cannot find the file specified.
it looks like it was previously able to create the service without admin, even though it was saying it didnt
Do I need to first launch the server not as a service and go through the initial setup in-game? I just cant for the life of me figure out why its giving me that error, when the wiki says nothing about that, and searching in this server also reveals I'm the only one who's asked about this haha
After the install call there should have been a window popping up to configure the service
yeah, it did that even without admin. although I noticed that in PowerShell it still said access denied despite me doing it. I assumed it wasnt working, but I guess it was.
doing it again, this time with admin, just says its already installed.
then I go to run it and get that error
Run "nssm remove <service name>" and redo with admin rights
if I create a .bat with the command in it and run that .bat as admin, that'll be the same thing right?
Depends
To simply start the server? Sure
To make it run as a service with background execution, autostart and auto restart? No
in my case, i was just thinking of running the .bat with "C:\Users\mini HP\Desktop\Satisfactory1.1Server\nssm.exe" start SatisfactoryServerService
No why? For one time just start the exe yourself. For autostart just set the service to auto start
I'm probably just misunderstanding what this nssm service thing is actually doing.
I assumed that by running that start SatisfactoryServerService its running FactoryServer.exe through that, and therefor will restart it automatically when it shuts down every 24 hours
NSSM just registeres something as a windows service
And the integrated daily restart is as the name implies a restart
it never turned back on for me when I was running a server in the past
it just stayed off
which is why im going down this rabbit hole
also, I guess I've never actually looked up the definition of a "windows service" specifically
I take it that means its registered in Windows as a thing that it must always keep on, even if it crashes.
Okay weird, normally it should
Previously I had installed it through Steam, running in an offline account on a separate PC. It was the lowest barrier to entry method, since I have absoltely no idea what I'm doing, clearly
So maybe that Steam version didnt auto restart
If you just want to keep restarting it just make a looping script that runs the exe
Registering it as a service allows you to run it without somebody logged-in (headless), and depending on the configuration auto start it on boot and restart it on exit/crash
not going to lie mate, considering I havent seen a step by step guide on doing that, I'm going to guess that's a lot tougher for me to pull off.
I have to admit that a few hours ago was my first time even opening Powershell. Let alone knowing what to do to even "navigate" in it.
This is why I asked about this earlier lol #dedicated-servers message
Never used it personally but look into WindowsGSM then
will do, thanks
and I have a followup dumb question
Shouldn't I be able to type in the cmd window that pops up when running FactoryServer.exe? like "stop" and "server.SaveGame saveName"
I thought I remember doing that last time I tinkered with server stuff
Idk
haha ok fair, thanks for the help thus far though, appreciate it.
I have been experiencing this issue with the last 2 build of 1.1 (That's the length of time I have been using it). In order to clear the problem, the process must be terminated and restarted (the internal daily "reset" does not fix the issue).
The problem typically escalates one core at a time until all cores are pegged at 100%. The server is a Ubuntu LXC container with 4 vCPUs and 8GB of RAM.
Another unsolicited post because someone might want this for their server when it releases.
Finally decided on how to have the cli command structure. Added some very basic commands for now. Have a prototype locally for server specific features without API. Have to figure out how to get these platform specific though for those who run on Windows servers not just linux.
Hello pioneers. Reddit led me here. Has anyone yet figured out a fix for the hologram bug with walls and foundations? Or are we just relogging forever? If it matters, it's Bisect 1.1 beta server, we told the website 20 people and took the recommendation despite only 6 of us in the end. We're in Chicago but had to host in Virginia as Chicago was sold out
Did not work for me
yes, thats a problem with port forwarding then
can he access https://<Your-external-IP>:7777 ?
yeah browser
then your port forwarding is not set up correctly
(i assume you use the default port of 7777? (+8888 for 1.1))
I did not know about the 8888. Guess I will need to look into it and the forwarding
@deep nebula Use canyouseeme dot org to test the Satisfactory ports. It will tell you if you're port forwarded correctly
Error: I could not see your service on on port (7777)
Reason: Connection timed out
guess portforwarding in deco app doesnt work
What is the deco plugged into? Is that modem/router/ONT in bridge mode? You could be facing double NAT meaning. Whatever the deco is plugged into is blocking the connection.
Whenever you bring your own router to your ISP connection you need to ensure the ISP is in bridge/transparent bridging mode so the tp link does all the firewall work and the ISP device simple hands off the connection
deco is plugged directly into the fiber converter and should be working as my router
Is that fiber "converter" a router or just an ONT? Some of them need to be set to bridge mode. I'd Google the model number and ensure it's set correctly if applicable
I guess it is an ONTπ€ but I will check
Is there a known fix to the issue of my dedicated server appearing as offline even though it's still running after I disconnect? I keep having to restart it after I leave. (I've so far been the only player on)
is anyone here using a hosting server they are happy with?
I'm having an issue with my server I'm hosting through GPortal, was working fine last night but now as soon as both my friend and I join at the same time we both get timed out and the server goes appears offline. Anyone found a fix for this issue?
Hey guys - I tried to updated my dedicated server to expirimental tonight using -beta experimental and the process seems to work fine but I'm now getting a "Version mismatch" in my client (also on beta). Saying my server version is 385279 and Game Version is 415558
"D:\SteamCMD\steamcmd.exe" +force_install_dir "C:\GameServers\SatisfactoryServer" +login anonymous +app_update 1690800 -beta experimental validate +quit
Known Issue? Searching here not seeing anything
server didnt update
You'll want to run the command again
Hey guys, how I modify the server max players?
I tried modifying the docker-compose file and wiped my sever
π
Who is the best/lag free host atm? I've tried in the past and it always has that little lag/delay.. Not worried if it's pricey just want something that works well
interesting question maybe idk
can i run the dedicated server on another linux computer (my home "server") if i dont have steam or epic games on it?
i mean pretty much just sending the steam downloaded files over to it and trying to run it
looks like ill have an interesting time running it as a service without having steamcmd but maybe i can set that up and do things properly
I self host from home and haven't rented a server in quite some time. If I was renting one I'd try indifferent broccoli and shockbyte side by side. Both have 2/3 day trial or refund windows. "Pricy" game server rentals are over $140 per month. At that price you could buy a MINISFORUM UN1290 for $391, own it and host multiple game servers with a couple clicks, blazing fast with no lag by installing amp on it.
You can, the dedicated server installation on Linux will install steam cmd.
How do i update my experimental server on Linux? We already tried to use app_update but did not work. Still get version mismatch
Take a look here: https://satisfactory.wiki.gg/wiki/Dedicated_servers
Specifically 3.2.3
Hey just wanted to see if anyone's had any issues with 1.1 dedicated servers with having issues placing stuff? Places but doesn't show it there when it is etc
did you double check +force_install_dir argument ?
Yeh we fixed it, it was that.
Good evening, I would like to play online. how can I do that
Do server restarts pause the game to the state before restart?
Well, yeah, I'm new to this, although not to play on Satisfactory, just the online part. I know how it works on Farming Simulator. But here I dunno so if there a group and we can play together it would be nice.
no public servers, no lobby, etc
find a group on that channel that you can help out, or post one for yourself. Servers and peer to peer both work about the same
!wikisearch multiplayer
Satisfactory can be played in multiplayer, with up to four players officially supported, but with no hard player-limit see below. The game is co-op only and there will be no PvP elements, although pioneers can damage each other. This means that HUB, MAM and Space Elevator progress as well as AWESOME...
Thanks everyone
so i bought a server maybe 5 months ago on shockbyte my friends and i played for a bit then stopped and i canceled it
how would i get that same save back if it was through the server?
It's gone unless you manually saved it.
You could log into your shockbyte panel and see if any of the backups exist. But I doubt it
They auto save then completely reboot. you get kicked, it comes back online then you can rejoin. Mine takes less than 30: seconds.
Online as in game clock continues? Only asking bc Iβm hoping to step away for a day or two and want to know if Iβll come back to full storage containers or not
If you have auto pause on in the settings it auto pauses when no one is on, turn it off and you continue to fill storages
Yes Iβve already done that and it continues after disconnecting, I was just wondering if server restart effects that progresss
I havent noticed it affect it, I have mine restart at like 4 am xD
Any news about this topic? Currently i monitor my cpu usage with a .sh and reboot my LXC Container when its over 50% cpu usage -.-
how long does it usually take for patches to get deployed to servers ? I restarted my server after the patch came out but it still says version mismatch
how do you start the server? mine starts per system.d and search for a update ever reboot... see https://satisfactory.wiki.gg/wiki/Dedicated_servers/Running_as_a_Service#Systemd
Mine runs on windows server 2016, same issue here
Anyone else having download issues?
steamcmd downloads the new patch with ~30KBs altho I have 1Gb/s bandwith
On Debian download was lightning fast with 1gb/s bandwitdh - maybe anyone gets the update now ^^
Dont know if it helps, but here is a BIG logfile from my Server since a few month. The Server was updated ~1-2d since experimental was released. Maybe you can get some information from it π
Hi, does anyone know how to fix a stuck in loading screen Error? The server uses my internal ip and the firewall only allows a specific ip adress that connects to my external ip. I host the server self on port 7777 but like i sair my friend is forwver stuck in a loading screen.
just made a ticket on the same question. My friend and i had over 400hrs in the same world and one day he just couldnt connect to it. Tried to give him my save file and he could open it but i could not connect. And the funny part is my second friend could on both of us
try opening port 8888/tcp too
And how do i enable this on the server?
You don't. It's a firewall rule.
You should have an allow rule for 7777 and 8888
If you don't have inbound allow rules. When a user goes to connect to your server. The firewall essentially says "No. I see you're trying to access a service on my network, but you cant connect because I don't have a rule letting you in"
how to update the dedicated server (steamcmd) install to 1.1 ?
If you restart it should update automatically.
Someone can't join their own dedication Satisfactory Server?
If I'm connecting to my server I have an infinity loading screen. But I hear that I'm doing something like shooting.
can anyone confirm this is correct version?
Build: ++FactoryGame+rel-main-1.1.0-CL-416835
yeah obv is even says 1.1... nvm
abit too eager to get going i guess
Open port 8888/TCP in the firewall
Is this a new rule after the update? If already opened 15000 tcp / udp
15000 is a pre 1.0 port and is no longer required
With 1.1 port 8888/TCP was added
alr. I will try it out.
This solved my issues, thanks
+1
"As of Patch 1.0, port numbers 15000 and 15777 are no longer used. At all. Sometimes port 8888 will need to be opened. " from the offical wiki ....
"sometimes"
I belive it should pull the newest version automatically, but i dodnt know, i have infinite loading
same for me
need the 8888 port or what ever you set it to if you get infinite loading
i already did that in previous patch (1.0)
if your in a docker container, needs to be added at system and docker level for adding of the port
services:
satisfactory-server:
container_name: 'game_satisfactory'
image: 'wolveix/satisfactory-server:latest'
volumes:
- './satisfactory-server:/config'
networks:
- traefik
environment:
- MAXPLAYERS=4
- PGID=1000
- PUID=1000
- STEAMBETA=false
deploy:
resources:
limits:
memory: 8G
reservations:
memory: 4G
labels:
- "traefik.enable=true"
- "traefik.tcp.routers.satisfactory-tcp.rule=HostSNI(`*`)"
- "traefik.tcp.routers.satisfactory-tcp.entrypoints=satisfactory-tcp,satisfactory-message"
- "traefik.udp.routers.satisfactory-udp.entrypoints=satisfactory-udp"
- "traefik.udp.services.satisfactory-udp.loadbalancer.server.port=7777"
satisfactory-tcp,satisfactory-message is 7777 tcp, and 8888 tcp, i also enabled it in ufw
Did you enable also UDP? That fixed it for me.
server update out yet?
Hello, does anyone else encounter the issue on dedicated servers which has been updated to 1.1 where players are stuck in infinite loading screens?
im not sure if it was the way my docker container did it, but it has both tcp and udp forwarded to the container for both ports
and it works?
yep
The image must be update so we can pull the new version
Has anyone found an updated docker image?
IΒ΄ve got the new update along with the update for the client. But IΒ΄m suing a Egg within pelican panel. idk if there is something other without pelican.
let me try that, one sec
I did. Same error.
im using pterodactyl, and the 1.0 egg, just need a small modification to work
Does dedicated servers work fine after the 1.1 release? Mine doesnt seem to work as it should.
I'm stuck too
mines not working
It does as long as you make sure ur server opens up 8888/tcp, as said above
Yeh I do
both are up to date but says mismatch
I got the update for the server to be 1.1 and it shows on server manager. But we are not able to connect. Infinite loading screen.
yes I did. Same error
i also have mismatch, dedicated server is not being updated properly with steamcmd
yeah seems so
And just to state the obv, its forwarded on router level and open on os level?
Mine doesnt show mismatch no more. But still not able to join in
what did you change?
i tried steam dedicated server and it works properly
yes
huh. Did you install it on a new server?
yeah mine says its up, I was just on it till I logged to update
yeah i reinstalled on steam i think, i used steamcmd for 1.0
added the 2nd port as a new allocation to the node, then added that port to the server, then added -ReliablePort=8888 to the startup command
opened 8888 on udp still nothing so that was not the problem
Any fixes for the infinte loading screen?
Got the same issue currently
What modification is that?
added the 2nd port as a new allocation to the node, then added that port to the server, then added -ReliablePort=8888 to the startup command
@spark fox this. But it may not apply to you if you're not using pterodactyl panel
I'm using Pelican, but it's close enough π
Change the TCP and UDP to 8888 it worked for us
Goat. Fixed it for me on pelican panel.
awesome
dows anyone use base wolveix/satisfactory-server and made it work?
hello, mine does this too. I have an error in my client console that says: ERROR: no area found please investigate. Did you find a fix for this?
Could we log file from the server of that session please ?
We are still in info collecting stage, its going to take a while to investigate, fix, test and figure out how to deploy the fix safely.
Already updated the server to 1.1
Also tried reinstalling it, and it just gets stuck for us
okok im getting there, the problem was somewhere in my traefik config, as when i enabled it with the basic
ports:
- '7777:7777/tcp'
- '7777:7777/udp'
- '8888:8888/tcp'
it worked, thats something
anyone know how to update gportal servers to 1.1 lol stuck with mismatch
There seems to be some issues right now
been playing around with the server, i can connect locally (my machine to access point to my Unifi Dream Machine Special Edition then the server is wired straight into there.
accessing via the external address a user connects and appears in the world, they can hear my actions (jumping etc) and i can see their interactions with mouse movement, but that appears to be all. the external user however remains stuck on the loading screen
open the secondary port, defaults to 8888
so does it never use the 7777 port anymore? Should I close it?
I run in docker, so, anyone fixed It?
server uses both 7777 and 8888 for ports now
gotcha
where were we supposed to find out about it π
i troubleshoot for two hours before checking in here
Should we connect through port 8888?
no
@mild thunder is ur server working ?
connect to 7777, and the server tells the client what the 2nd port is
my steam server wont update did the comman on steamcmd says it downloaded and verifed files but when i connect tells me steam client mismatch to server
yes
oh ok. but what did u unnormaly π i cant patch it π
services:
satisfactory-server:
container_name: 'satisfactory-server'
hostname: 'satisfactory-server'
image: 'wolveix/satisfactory-server:latest'
ports:
- '7777:7777/tcp'
- '7777:7777/udp'
- '8888:8888/tcp'
- '8888:8888/udp'
volumes:
- './satisfactory-server:/config'
environment:
- MAXPLAYERS=6
- PGID=1000
- PUID=1000
- STEAMBETA=false
restart: unless-stopped
deploy:
resources:
limits:
memory: 8G
reservations:
memory: 4G
i stopped my server, told it to reinstall and then started it after modifying a few things
yup me too.
will give this a try
so u dont update it u reinstall it ?
This is my docker compose, anything else I should do?
im using pterodactyl, and the quickest way for me to force an update was to click the reinstall button
yes, this is my bat script as an example
.\steam\steamcmd.exe +force_install_dir "D:\satisfactory_server\server" +login anonymous +app_update 1690800 validate +quit```
i tryd this a couple of times π
opening up and forwarding 8888 doesn't seem to change anything
were you on experimental before?
nope. never on experimental
I updated my server using steamcmd and my saves are gone. Is this normal behaviour?
Did you make a backup?
bat file, you hosting on windows?
yeah
yes
Thank god, because i did not First time it happened
Does anyone have a fix for the problem of being stuck on the loading screen since the update?
watch the server startup, and see what ports it says its using, to verify that it can allocate the proper ports
I can send you logs If you are available
i can try to take a quick look, trying to get stuff ready to move to the office as i worked from home this morning
did changing the port in a script file get broken? I needed to change my port due to other servers being hosted but the server is still trying to bind to 8888 even though i have it forced to a different one in a .sh file
For the Pelican people, a change is on the way, and the egg just needs to be updated:
https://github.com/pelican-eggs/games-steamcmd/pull/159
https://github.com/pelican-eggs/games-steamcmd/tree/main/satisfactory
@mild thunder ehmm... now it works after i reinstall the game. idk why and dont know the "because" but. It works. Sorry cant help anyone idk why i works now @ me >_>
need these both as the server takes 2 ports now -Port=7777 -ReliablePort=8888
Anyone else stuck here with their server?
Multiple people are
Hey ho Pioneers, happy patch day o/
Probably has been asked a dozen times, but is there a way to click join server while having a Controller connected to the Steam Deck? Or a Controller in general.
The dedicated server host needs to open their port 8888 on TCP
Already have that. Just using different ports
Well the game port needs to be open on TCP/USD, and the reliable port needs to be open on TCP
(Also, how do even force different port? Docker container?)
Yeah. I can also run multiple instances like that
yeah, that's fair
i wish there was a way to manually host multiple instances, instead of it relying on appdata.
could be minecraft like, with its own data folder in the dedicated server folder, and a config file to setup ports. way easier than a docker container
I can recommend Pelican. It also gives a nice interface to handle it from π
Never heard of it but it does look really cool
Some of the guys that made Pterodactyl split, and made Pelican, based on Pterodactyl
I just needed to open TCP 8888 on my firewall at the server and after that it worked fine after update
is this a new required port?
i've tried opening (and forwarding)8888 and still having infinate loading
what else could be?
Did you open for incoming connections in firewall?
I opened both 7777 and 8888 in docker compose file and it's now working
i already fixed it was my traefik config that was messing up not the satiasfactory server
For everyone? I can load fine but my friends cant
haven't had others try yet, can update you later. make sure your external firewall also has this opened and properly forwarded
how do i check that in a ubuntu server enviroment?
i use my public domain address to connect to my local server and it works, so I assume it's a firewall issue for you
also need to check your router firewall to make sure ports are forwarded to that host
yea i can connect locally to my server, but while people can connect to and join the game remotely, they don't get past the loading screen (can see them in game and they can hear everything around and 'look around')
so they're definatly connecting as i can see them interacting in the level
but they just have the loading screen still up
That was the first thing I did, honestly, and I made sure in the past that I have a static local address as well
tcp6 0 0 :::8888 :::* LISTEN 4166710/docker-prox
open 8888/tcp on your server, and let them through your router if server is local
same for you i guess, check the router
have done so
It is open π
It's not the same for everyone. I already TCP forwarded
i dont know what firewall you use on your server but check if its open there too
how I can use custom port configs?
i don't know if it's necessary or if you did it, but I did compose-down and compose-up to recreate the container
I did, but, will try again
yes that has worked, forgot that virgin hub needed port forwarding enabled aswell (idiot)
keep forgetting that its not just a pass through to my actual hub
Update 1.1 Roadmap:
- if dockerised enable port 8888/tcp (dont forget to restart the container)
- on the server machine allow 8888/tcp on firewall (for ufw: sudo ufw allow 8888)
- on router (if server is local and players are not) enable portforward 8888/tcp to your server's ip
- if there is a middleware (traefik, nginx, etc... ) configure it as needed
If I missed something please respond!
let me try to help future misguided souls
indeed i am
I've a serious question.... is the pc getting hotter now than bf the update?
I am so lost. I got the update to work, finally. (I had to move the original files to a different directory and reinstall, then moved my savefile folder back). I can see the server in the server panel in the game. (I can even send the exit command via terminal and it works.) But I cannot connect to the server.
I am not well versed in Linux (using Ubuntu).
So far i've tried (on docker):
- Put the new ports in, both TCP and UDP 8888
- Portfowarding in router 8888 both TCP and UDP
- sudo ufw allow 8888 (Which gave me the message below, even though I believe that ufw is not active atm)
Rules updated (v6)```
if you telnet from outside do you get response?
Using Ptero or only Docker?
only docker compose from https://github.com/wolveix/satisfactory-server
Me? I've shut down the server at the moment. I haven't tried telnet. I haven't used THAT since the 90's.
you can check as well, its still great if you want to check if port is actually open on the server machine
I may have to wait till my partner gets home so he can set up the 8888 port on the router.
Both ports open in same Docker?
services:
satisfactory-server:
container_name: 'satisfactory-server'
hostname: 'satisfactory-server'
image: 'wolveix/satisfactory-server:latest'
ports:
- '7777:7777/tcp'
- '7777:7777/udp'
- '8888:8888/tcp'
- '8888:8888/udp'
volumes:
- './satisfactory-server:/config'
environment:
- MAXPLAYERS=6
- PGID=1000
- PUID=1000
- STEAMBETA=false
restart: unless-stopped
deploy:
resources:
limits:
memory: 8G
reservations:
memory: 4G```
sudo netstat -ntlp
both ports listen?
tcp6 0 0 :::8888 :::* LISTEN 20316/docker-proxy
tcp 0 0 0.0.0.0:8888 0.0.0.0:* LISTEN 20307/docker-proxy
tcp 0 0 0.0.0.0:7777 0.0.0.0:* LISTEN 20276/docker-proxy
tcp6 0 0 :::7777 :::* LISTEN 20284/docker-proxy```
yep
only 1 server?
yea
only one instance running ?
yea
if behind cgnat cannot host any workaround
tunneling
It was working normally until 1.1
telnet <ip> 8888 what does it return
public ip?
whats this?
ye, so you check for port forwarding
a docker image
first tried this:
Server lookup failure: (REDACTED):telnet, Name or service not known```
Trying (REDACTED)...
Connected to (REDACTED).
Escape character is '^]'.```
tried both ways
these were the results
by both ways what do you mean
oh i see
why telnet ?
so its reaching
why not, does the job
do you use any reversy proxy
ok I didnt know it was installed by default
nope
its just a homelab
for this kind of stuff
any good hosting services
for the time being remove the ports and try network_mode: host
going to https://<server_address>:7777/ in a browser usually works well to check if the server is up
u might be behind cgnat also if not trying to connect locally
@pseudo ravine I see in your firewall that you have :8888 configured for tcp. It needs to be configured for udp
No
i dont even have 8888/udp open and it works fine
the docker image doesnt even listen on 8888/udp
{"errorCode": "errors.com.epicgames.httpserver.route_handler_not_found", "errorMessage": ""}
π
thats good, it just show its open
changed and recreate the container, still like this
i was hoping for STATS
I can listen to the game, and sounds
ye its the 8888 thing, had the same for me untill i fixed it
I changed ISPs now to try the remote connection
And that is the error
I can listen to the game, but it does not complete the loading
The weird thing is: I do not know where else to port forward
the router already is configured correctly
Am I missing anything in the docker image perhaps?
how I can config custom ports for satisfactory
i dont think so, its kinda the same for me
So, its not working for you too?
it does, thats why im trying to help
π
im already out of the blue with this one
We need a network wizard
have you tried restarting everything, your router, your server btw?
No
I have this issue.
if I move my mouse I can hear the sound shifting around as if my pioneer is moving
but the loading screen won't frickin' go away
I doubt it's network related because the dedicated server I'm on is hosted on a server that's racked 3ft to my right and I have a friend who can access it no problems...
Just trying to verify the game integrity to see if this fixes it - will keep you posted.
my dedicated server is also in the same room as my pc, I can access it locally, my friend cant
its a portforward thing for sure then, have you tried restarting your router
funny that
Mine's the other way around - I can't access it locally, but my friend, who's in the Netherlands can...
thats some other funny business
im not reallly new on hosting server but using the dockers do i need ubuntu?
no, any os is great if you have docker daemon
pi aswell? xD
The thing is, what else can I try?
ask chatgpt i guess
i prefer to die
ended up hosting a server on hosting services, because im on behind cgnat
maybe in your router config the port is block by routers firewall, no clue
Noticed this error:
[2025.06.10-18.53.58:474][556]LogAkAudio: Warning: UAkGameplayStatics::StopActor: Could not retrieve audio device.
[2025.06.10-18.54.00:023][602]LogReliableMessaging: Error: Received an invalid message type. Disconnecting.
[2025.06.10-18.54.00:026][603]LogReliableMessaging: Incoming TCP connection broken before it could be established.````
thats something
In docker logs