#dedicated-servers
1 messages · Page 10 of 1
ok when i do the whole log -unattended my satisfactory game count the server as offline :d
@vast sigil What does it say when you try to launch it?
uh nothing really it launchesm i go in to check server load to log in aaand "server offline"
how do you start it?
whats in the bat
factoryserver.exe -log -unattended
your cmd.exe which will run the batfile might not find factoryserver.exe if it's not in the PATH environment variable, and you may have to provide full path or add it to PATH
nothing in logs about authenticating
Is the server on a dedicated server or your PC?
u asking me or?
dedicated
Have you allowed the port through firewall?
firewall is off for test
@vast sigil Did you try start->type cmd -> press enter, then type the command? What does it say?
Hmm, i dont know - I never had any issues with authentication so i'm not sure from here
Where does the passwords store when you set them
i can pull the game status by entering player password i set from before i wiped which is odd
i totally deleted everything including steam cmd folders
im in command console but rarely use it
redownloaded everything and re validated and downloaded dedicated server
Ok, then write full absolute path of the factoryserver.exe into your batfile
then it will find it
or, add the directory where the factoryserver.exe is into your PATH. You can do that by pressing start->write environ and you should get an option to edit environment variables
PATH environment variable is used to look for .exe files to use as cmd commands
is there a way to find out what patch on EA game and what is dedicated server patch?
EA= early access or live
can the address part not be a url like play.signaturedom.com ?
@spice heart no
Idk ask the devs lol
@spice heart that's incorrect, it's an ip, and as such you can have whatever public ip or dns name you want, just like an Mc server.
@desert walrus really?
I plan to use play.satisfactory.dev lol
Weird, I've joined several servers using something.something.co
this not look normal to you?
Mabe he did something wrong with the domain
yeh its been claimed
Ok, so you were able to connect to it over lan?
Also, you're sure that everything with your domain is set up properly and it's liked to your public ip?
Does it work if you enter the ip, instead of a DNS name?
It's just the ip of whatever it's hosted on
If it's on your machine, just Google "what's my ip"
You could also try this command, but you shouldn't have to unless there's multiple things hosted on your machine.
Also make sure your query port is correct
Hello guys, is it normal that right after launching the server, it is instantly shutting itself down?
The server keeps the server configuration (passwords, session name, server name, various other shit I'm not sure about) in the user's home directory.
It's a binary file, and the passwords look hashed at least, so I doubt they can be recovered. You can, however, remove the file to start over.
Or, you can just change the server's Query port, because that's what is used in the filename (by default the port is 15777, so the default filename is ServerSettings.15777).
What address did you use?
Ah, it's likely that the DNS record doesn't exist, and that's why it said it doesn't seem to be valid. That's a bullshit error message that doesn't tell you what happened. It's user-friendly, but completely useless. xD
I get a laugh each time someone tries to tell me that error messages are "unfriendly" or tries to make them "friendlier". :D That's so much bullshit.
The error message should plainly state "The domain name you entered does not resolve to an IP address." or something. That's likely the most informative. Alternatively, "The domain name you entered does not exist."
"does not seem valid" encompasses:
- actually invalid name (i.e. non-compliant with the RFC)
- tld/domain/subdomain doesn't exist
- subdomain doesn't have an A/AAAA record
- no DNS server responded to the query
Hey guys, I cant find the section pertaining to the beta version of the dedicated server. I'm running the dedicated server on Redhat 8.6, and when I open a drop pod. there is no icon for the hard drive, though I can still double click on it and recieve the drive
12 cores, 28G RAM on 4xSSD's set RAID 10. It's running in my network, so there is no latency, and resources appear fine on the server. No high I/O, CPU, or memory usage
Latest beta release of the linux version of the dedicated server
The dedicated server is beta for all intents and purposes. The missing Hard Drive icon is a known bug. The item is there, and if you double-click it - it will be placed into your inventory.
Ah, good to know. 🙂
Similarly, if you close the pod and reopen it - it'll load up correctly.
> so there is no latency
Hah. Unreal Engine was the first time I found out that you indeed can get rubber-banding on a LAN server. 🤣
ping is like 5ms :p
That's pretty high for LAN, actually.
Running a 6 Server VMWare Farm, interconnected with a 24 port 1GB switch. That's over my wifi on the other side of my house :p
Ah, the WiFi does you in then.
I'm seeing avg 1.75ms over VPN to my workplace, but that's all Cat6 and fiber.
Nice! My server room is downstairs in the lower bathrooms closet, dont want to sit on the toilet and play hardwared into the network :p
I tried playing on the toilet once. A bit unwieldy. xD
lulz xD
I mean, the keyboard and mouse need some space, you know,
and hanging even an LCD next to the toilet paper is... well... xD
Just make sure your finger doesnt go through the TP. It'll be all downhill from there :p
TP nothing, man. Just make sure your finger doesn't get dirty, or that you don't accidentally touch the power button on the display.
Those don't wash easily.
lol
I assume you're playing on something like the Switch, or Steam Deck.
Still, even if you shit for 20 minutes ||like I do||, you could just wait.
... uhhh, no? 🤣 But are you sure it's actually shutting itself down? On Windows for example, if you don't add -log to the launch parameters - it's going to immediately hide the console window for some reason, so it might look like it's shutting down, but it in fact continues to run.
Nah I just smashed my keyboard and it launched itself
I don't understand what happened
It seems to have fixed itself
If you mean that it "stops and does nothing" and you're on Windows, make sure you don't accidentally enter Select Mode.
(you leave it by hitting ENTER)
Hmmm
Speaking of which, is there a log file for the server? Just started using it a few days ago, haven't read into the command line or config options yet
There is. You should find the logs in /FactoryGame/Saved/Logs.
nice
The latest is FactoryGame.log, the second latest is FactoryGame_2.log IIRC.
Older logs should have timestamps.
Dated by launch time.
NOTE that the logs keep UTC timestamps. :/
Would setting an auto-rotate via cron hurt anything?
It auto-rotates on launch.
err, I meant autocompression
And as far as I'm aware, it keeps the handles open.
Probably not, as long as you don't try to touch the currently-open log, or the _2 log.
Like, compress the old logs after x amount of days, move them to an archive
okies
perfect, some applications are touchy with their previous log files
You can essentially compress anything named FactoryGame-backup-DATETIME.log
since that's been renamed by the server and shouldn't be touched any further.
Cheers, found it in %appdata% local
Huh, you can even one-line it. su - steam -c "cd /var/NVMe/SATISFACTORY_Dedicated_Server/FactoryGame/Saved/Logs; find . -iname \"FactoryGame-backup-*.log\" -exec gzip {} \;" works like a charm. :D
Love it, nice and easy :p
Also, if you see 'CRC' in the file, it's not a CRC32. xD
lol
CRC stands for Crash Report Client. 😂
okies
Are there a lot of people that use the linux server compared to windows? Just curious about the ratio. It was so nice to see it have a linux server package.
No idea. I doubt there are any public statistics.
I can see that it's pretty much 50/50 from where I'm sitting. I have six UE games (ARK, Satisfactory, Memories of Mars, ATLAS, Conan: Exiles, DnL), and three provide a Linux server (ARK, SF, Memories of Mars). DnL works well enough in wine, and Conan: Exiles used to, but it no longer works as of the latest update. ATLAS only really works on Windows.
So yes, I'm damn happy to have a Linux binary for Satisfactory DS.
I am trying to make a dedicated server using pebblehost. Every time i load up the server and try loading a world I get a timed out and the server crashes. Do you guys have any advice?
Post the log to really see what going on but if it crashes when you try to log in your probably running out of ram as the new update you need at least 10gb free to be able to load the map
Oh ok thanks, will look into it
Heya folks. I'm trying to set up a dedi server, I have my ports forwarded to my local IP on this machine, I've set up the server, connected to it locally, started a new level, can play, etc, but none of my friends can join me.
I checked through powershell, it has all the local address ports and processes listed what looks correct to me, and I've tried with and without -multihome, using either 0.0.0.0 or my LAN ip for this machine.
Running out of ideas and troubleshooting steps on the wiki
Oh I also messed with network quality Ultra, logging out of epic games account (Running it on steam), and disabling "Send gameplay data"
if you can connect it from your local network, but your friends can't, then it means your port forwards aren't correct
if you're running it on same machine as you play on, then it could be firewall?
defender fire walls shows as "Allow" for private/public for two satisfactory apps, and satisfactory dedicated server
also, you did forward udp or udp also?
i don't know about defender, is that for outbound or inbound stuff? My guess is firewall or protocol (i don't know what satisfactory uses, but at least udp on some of ports)
i don't know how to use. I can click things and they start on windows
Well. That's where my issue is, everything looks correct
ports are forwarded, firewall is allowing the apps through
no dice anyways
someone might wanna read the backlog and ask more questions or tell where you might have gone wrong 😦 I don't know.
I've set up dedi servers in other games before, so this isn't my first rodeo
just seems to simply not work
I did recently get this fancy GT-AX11000 "gaming" router with a ton of features I don't really know about, and am worried that might be the cause
but I've not enabled anything specifically on it, and disabled IPv6 firewall
Also kind of weird: When I connect to my own server, it alters me between two player profiles it seems? So I'm just cloning myself
These are the ports I have open on my firewall on my server. If you didnt change the config on the server, just mirror these ports in your router or edge firewall, and you should be good
[steam@nixgame ~]$ sudo firewall-cmd --list-ports
7777/udp 15000/udp 15777/udp
[steam@nixgame ~]$
and you are giving your friends your public IP to join correct... and btw you dont have to hide you internal ip lol
Heh, last time I was told off for not so
but yeah without my public ip matters little
and yes
Alright. I changed my NAT type in my router to full cone from symmetric, and they were able to add my server in manage servers and join me
those are people who dont know about networking lol but yeah there is no need to hide your internal ip
It may be good to make a post, stating what type of edge device you run, and the modifications to NAT that was needed to make it operational externally. Might get posted in the FAQ's to help other people.
If there's any mention of Killed in there, you're running out of RAM. Increase allocation to 10 GiB.
Huh, I was not even aware that this was a thing... Live and learn, I suppose.
However, a full cone NAT should work for Satisfactory. A symmetric NAT is more restricted than full-cone.
It should be mentioned that a typical NAT implementation is a mix of the two - most communication is symmetric, but port-forwards create channels that behave like full-cone for those respective ports.
Symmetric is usually achieved by doing something like iptables -A FORWARD -m conntrack --ctstate RELATED,ESTABLISHED -j ACCEPT, which automatically accepts packets that are coming back on a channel that's already been created by outbound packets.
(conntrack is the module that keeps all of these in memory)
Packet filters (and I refuse to call these "firewalls" :D) work in a very simple way - they read the rules, and go in a loop:
for each (rule IN ruleset)
if (traffic MATCHES rule) APPLY rule
// at this point we've cycled through all rules in the ruleset; if a rule has not terminated (or if no rule was matched), we apply the default behaviour instead
apply DEFAULT
A rule can either terminate, or return. If a rule terminated - the filter reached a final decision on what to do with that traffic. If the rule returned - the filter will continue through the rest of the rules until termination. The DEFAULT rule always terminates.
That's about it.
hello i'm french sry if my message isn't right
on my dedicated server i have a problem, when i'm hit by Spitter my player move on right and left really fast ..
2nd problem when i use gun with auto bullet they always shoot on my legs, so i need to jump high to make sure hit mob
sur mon serveur dédié j'ai un probleme, quand je me fais toucher par un slitter mon perso bouge de droite a gauche super vite ce qui m'empche de bouger et de le toucher
seconde probleme quand j'utilise le fusil avec les munition auto la balle par sur mes pieds, donc je suis obligé de sauter super haut pour etre sur de le toucher
when i'm hit by Spitter my player move on right and left really fast
Known issue. Wait for developers to fix it.
auto bullet they always shoot on my legs, so i need to jump high to make sure hit mob
I also noticed this. Also needs a fix from the developers.
But hey, if it's any consolation -- if you hit them with a bullet, they get jerked around too! 🤣
Hello, me and my friend are having issues with extreme amounts of ping on my dedicated server
like over 700ms
over lan
What's your actual ping to the machine?
Also, tell us more about the machine in question - Linux/Windows, for starters?
the host is linux, and the direct ping to the server is 232
232 ms over lan?
yes
What are you using, null modems?
How big is this "lan"?
For comparison purposes, my ping to my server - over LAN - is a little under 2 ms.
So what do you have that is making it over 200x that?
I think scuffed lan, reconnecting the server dropped it from 232 to 1.3ms...
Is your "LAN" over a VPN?
no
yes directly, in satisfactory its 60
That's weird. What else do you have in that network of yours, or on that server?
Because I was actually mistaken. The ping I mentioned - a little under 2 ms - is from my LAN to my work office over VPN.
The actual ping to my server is under 1 ms, and Satisfactory sometimes displays it as 0ms.
I shit you not.
Are you doing a lot of P2P stuff? Do you have 100 Mbps segments inbetween?
What can be done about the desync between clients on a dedicated server?
Faster hardware
What sort of desync are we talking about, here?
doubt its HW related but thanks.
What sort of desync?
For example, rubber banding at intervals that everybody sees and experiences at the similar times but not close enough to be a "hiccup" of the servers network, its a 1gb connection with only satisfactory and SSH listening.. Other types of "desync" we are experiencing, are items moving along belts seemingly being present, until you try to collect them off the belt, the item vanishes from the belt, but never appears in your inventory, as if it was a ghost.
the item vanishes from the belt, but never appears in your inventory, as if it was a ghost
Ah, so the predictive algorithm finally strikes a miss. I was wondering when that was going to show up...
As to that: there's nothing to be done. I doubt this can even be fixed.
Moving on.
rubber banding at intervals that everybody sees and experiences at the similar times
Does this happen exactly at 5-minute intervals?
Havent timed it honestly, but if I were to guess that would sound about right, ish. maybe.
Then the answer is faster hardware.
What you're experiencing is "hiccups" caused by the server saving the game.
While it does this, it does not handle any game or network events.
Thus - rubber-banding.
Further: if it does this for longer than 30 seconds, you'll start experiencing timeout disconnects.
You can resolve this by increasing the timeout (both client and server), but that has a point where it becomes unplayable.
You can increase the auto-save interval, which will spread these 'hiccups' apart.
But ultimately, the only way to fix this is faster CPU speeds.
How
ok thats updated, but what about the ghost items on belts?
That has to do with how the game replicates items between clients.
Much like with client movement, it does predictions.
Sometimes, those predictions are inaccurate.
And what about Schrodinger's slugs?
You never mentioned that before.
Yeah I didnt get to it lol
Slugs that are "present" that have been collected previously
and are present at range, until you approach then they just vanish
And this is unpreventable as the server auto restarts what, every 24 hours? built in.
Beyond that, the game just shows the defaults, as it doesn't have the info on it. This is actually your client doing it, not the server.
I believe you can configure that as well, but I'm not sure if it works, and if you can set it to 0.
Still, the server can still crash from time to time... ;]
Anyone know a six for DXGI_ERROR_DEVICE_REMOVED with Reason: DXGI_ERROR_DEVICE_HUNG Its ONLY happening in multiplayer on a dedicated server
its hapenning to one of my buddies, and can't figure it out
that seems like his GPU crash
Yea i think driver.. but it's only happening in multiplayer
Should also happen on single-player, because that's GPU driver crash-related.
he says hasn't crashed once in SP
Which is not evidence that it can't. I would need to see the frequency of the crashes vs frequency and time of playing attempts to determine how much weight to assign to this.
It is entirely possible that what the server is doing (differently from SP) is creating situations where these crashes may happen,
but ultimately the fix is: fix your GPU driver or demand your GPU manufacturer to do it.
Game developers are not in the business of fixing GPU drivers. Historical evidence notwithstanding. xD
hmm thanks
is 16g enough ram for a few weeks?
It should be, considering that the game should - in sane situations - require between 8 and 10 GiB.
but 8.8 is apparently the minimum for a fresh game 🤣
Does loading a dedicated server always take a really long time or is it doing something for one time only? Like converting the map or something
I have been waiting for about 20 minutes now and I am not sure what's going on
only takes about a minute for me
Did you add a custom save?
Yeah, uploaded one. But it's pretty new.
Hmm, same here, also pretty new
12GB of RAM on an SSD should be pretty capable of running
Yeah but just ram size is one thing there is also processor speed and ram speed that matter how fast thing get loaded into the ram
Processor is steady at around 24% used
On a 3950x and 64gb 3466 ram kit load up in a couple of minutes for me
Once every 5 minutes I get this repeated
So it should be running, but it's not hah
have you tried restarting your client? I find that the server manager UI is a bit weird about refreshing
If your seeing that yeah that usually means is running
Ooh wait a minute, it seems like the server list isn't updating it and you have to re-add the server again?
no, it just sucks at updating
you can keep it added, but just like, exit the menu or client and open the game again and see if it's any better
I had a lot of trouble with that at first
Yea, doesn't seem to be working so well then
ture is driving me nuts and I demand that be fixed
Haven't seen that one yet but I argee
Are you running via systemd? If so, are you sure it isn't rebooting after the process crashes?
Took me a very long time to realize that was happening to me when there wasn't enough RAM
Nah I got it running in ptero, so it did indeed crash, but it said the RAM was too low
I'm not even done with tier 1 and this is me:
USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
steam 20 0 12.4g 8.9g 157092 R 27.7 57.2 27:23.64 UE4Server-Linux
And laggy or not?
its great, but I have 16GB in this VM
Yea, I might aswell set it to 16GB later then, first I'll see how it goes
I have 12 threads with 28G RAM on a Xeon E5 series proc on 4x SSD's set RAID 10, runs like a champ. I initially had 12G on the VM, but I moved it to 28 after update 6 was released
Best to have that cushion. Im on tier 5/6 and Im using
16878 steam 20 0 14.1g 10.1g 173552 S 25.0 37.2 863:45.80 UE4Server-Linux
Can I load a mod into dedicated server? I have searched, check the wiki, cant find anything
8.01 GiB, actually
If it's saving the game, it is running. There's no other reason it would be saving.
It sometimes gets stuck in this situation where it's "pretty confident" that it's not running - and nothing you do will convince it otherwise.
There's also flase, so I think it's on purpose.
I guess they wanted to see how many people would report it? 😂
Might be a small social experiment they're running on the side? Who knows. 🤷
Yeah, I'm at 9.4 GiB in the middle of Phase 3.
Not bad lol.
ah, htop, love that package :p
Well, it's nice-looking.
htop = fun to use
Shows the same info, so meh, idgaf. :D
but yeah. @blazing halo possibly if you can, do a restart on the game
see if it just has a "memory" leak or something
I use mcedit instead of vi too. xD
lol
I grew up on debian. Though I use vi now, I gummed my teeth on nano xD
Here is what it looks like on Tier 5/6
Iostat for it also
[steam@nixgame ~]$ iostat
Linux 4.18.0-372.26.1.el8_6.x86_64 (nixgame.stillevande.local) 09/24/2022 x86_64 (12 CPU)
avg-cpu: %user %nice %system %iowait %steal %idle
1.47 0.31 0.64 0.16 0.00 97.42
Device tps kB_read/s kB_wrtn/s kB_read kB_wrtn
sda 26.44 6407.96 270.74 3777815 159615
scd0 0.07 1.77 0.00 1041 0
Interestingly enough, the actual VM usage is quite small. Application reservation is quite large.
Holy crap, I just noticed it's RHEL. 😂 I don't usually have to interact with it, and SELinux is a major turn-off for me. :D
Especially for private stuff that doesn't really see the Internet.
Anyone else having crashes since Update 6? Still on experimental Dedicated Server. The process just terminates with 3 lines
[2022.09.25-11.08.47:982][ 31]LogWindows: FPlatformMisc::RequestExit(1)
[2022.09.25-11.08.47:982][ 31]LogCore: Engine exit requested (reason: Win RequestExit)
[2022.09.25-11.08.47:990][ 31]Log file closed, 09/25/22 12:08:47
How much RAM do you have available? Did you close the CMD window?
Didn't close the window, it is just gone when I tab and check. 24GB free on the box running it
Okay, I think the reason for the shutdown is given earlier. As far as I can tell, UE does some housekeeping/cleanup on exit, and then reiterates the shutdown reason before closing. Can you hand over some more of the log? Preferably, the whole thing?
@blazing halo Sure, can I upload it here or DM or what?
Whichever you prefer.
Yeah, I have no fucking clue what's going on there.
It's possible you have a fucked save, to be honest.
Looks like it craps out a lot on meshes and the customiser, and I recall there being issues version-to-version with painting stuff... and recently, the game crashes with specific foundation placements, supposedly. I don't know.
Load the save into local play, and see if it does the same. If it crashes, it might give more info (and it would be good if you could put the save up on the QA site). If it loads and works, try saving the game with a new name and loading that into the server.
You can also try loading it into SCIM, and exporting from there as a new save, then loading that.
Sometimes reading and rewriting the save in a different tool works wonders for data integrity.
I can't connect anymore to the server...
Did the server crash?
Yeah, that tends to happen sometimes... not sure why. Usually disabling the internal router helps, but I don't really know why that happens.
the server loads the assets and then peng nothing 😄
This is the savegame process - looks normal.
Anyway, I can't see straight; woke up 18 hours ago. Maybe someone else can help, because I'm out of juice for today. :D
Though I have to ask: do you have the server configured to pause when there are 0 players?
Because if you do, it's likely that the first connection after pause (or startup) will time out.
no it runs even with no one connected
Also, if you have enabled the option to save after a player exits - remember to wait 30 seconds after being kicked/timed out, as even a failed connection (if you see the loading screen) still counts as player join & quit, so it will save the game afterwards.
(trying to join while it's saving the game will likely time out, which will trigger the save again, so it's best to wait half a minute)
thats not the problem. I was gone for like an hour and now nothing works.
Just wanted to mention it. It's a fairly recent find/realisation.
Server is running on aws with docker and everything should be fine
I'll try redownloading the server, should do the trick, if the savegame isn't broken 😄
hi there!
i got some trouble while trying to launch dedicated server as service someone can help me please?
whats going on 😄
@foggy agate heeehhhh don't know? lol
i paste the file as shown on the wiki, replace the coresponding field to my account/path and... service won't run because exit-code
I get the error:
Steam API failed to initialize! then steam shuts down :/
Warning: Failed to read file '/home/steam/.config/Epic/FactoryGame/Saved/SaveGames/ServerSettings.15777' error.
so.. tried a new server, but still loading the savegame is ok but connecting to the server (when it acutually lodading the world) it runs in a time out.
so there needs to be a bug within the server
When the server is loading the map, it does not send any data to clients, if that takes longer than 30 seconds, the client gets a timeout. It's not so much a bug, just that the server performance is too slow to complete within 30 seconds.
makes sense, but still sounds like a bug.
If your server is faster, it wont happen.
its allready on the fastest single core cpu aws provides with gp3 ssd with 125 mbit transfer speed.
I have an old server with a Xeon D with SSD and it takes 20-25 seconds so I dont have the problem. Just for reference.
In update 5, the loading was faster so its more rare to find this, but it works the same. Update 6 just needs better hardware.
2nd generation Intel Xeon Scalable processors with up to 4.5 GHz frequency
thats the fastest cpu aws has to offer and when even this is "too slow" you could just switch to a desktop cpu setup with hetzner or something like that
Coffee Stain can do multiple things, either state higher hardware requirements or they can increase the timeout for clients to avoid this.
Well maybe something else is the bottleneck then, cause for me it takes 20-25 seconds to load on a 2.0Ghz cpu
Or is your save extremely large?
no, I'm on stage 4 right now, wouldn't call it extremly large
I tried the same save file on a dedicated instance on gcp, with the same result. Normally we use this for in memory databases, so there can't be any bottle neck besides the game itself.
this
and this
I doubt that this is the error, because I don't have to wait 30 sec. At least something has to be loaded, because if I load a savegame from the server locally progress on the space elevator has been made.
I don't understand what you are trying to say. On the client, when you connect to a server, there is a timeout for 30 seconds. That's literally what your screenshot shows.
When the server loads a map/save, this takes time. During this time, the server will not communicate with any client that is connected or tries to connect.
If the loading takes longer than the 30 seconds, your client will disconnect on timeout.
If loading takes less than 30 seconds, the client will connect as expected.
In case of the longer loading, if you wait till the server is done and then connect, you will connect as expected.
It all depends on the performance of the server hardware and the size of the savefile if you run into this problem or not.
That's network quality related as far as I'm aware. But a quick Google also shows network quality settings on server and client being too far apart and/or running broken mods and/or too many crates on floors etc.
Hello 😀
Just tried my server for the first time since U6 and it seems to be crashing while creating a new game.
First thing to try, reinstall the server?
just delete the STEAMAPPS folder inside the dedicated server directory, and re-run the steamCMD update. It should overwrite files
I'm trying to join a Steam Friend Session, i get "Outgoing Reliable buffer Overflow".. and get kicked.. What is this?
I thought you couldn't run MODs on a dedicated server?
I have somehow managed to mess my satisfactory DS up more, because now it seems to be crashing a minute after booting 😦
Well, it's not so much as that it's not possible, it's just that nobody has done it yet. It's most definitely possible. So I don't really know what people try to hack together these days and mentioned it anyway.
Hello all, how i can run with root user DedicatedServer?
Beacons broken in multiplayer?
if you dont have more around 10Gb free the server wont start up update 6 need more ram to be able to start up than update 5
ah, I only have 8, thank you!
there is no MOD support in dedicated server as of right now
yeah minimum requirements for dedicated is 12GB of ram and 16GB is recommended as stated in the WIKI
Someone?
what you mean by root user??? all you can do is use the in-game server manager panel and put a password for the setting that pretty much it there is no admin panel or things like that
When i try start in docker container as root user the DedicatetServer StatisFactory - server does not start and say me "Refusing to run with the root privileges."
yeah that a question for @blazing halo sorry i don't run it in docker so i wouldn't be able to help you there... try googling it maybe
I really try =))
I keep getting that error when attempting to log into our server... can work just fine, or take 10 attempts, or 2... it's entirely random it seems
I'm not an expert, but from my troubleshooting today, the standard first question is "Is your server powerful enough?"
If the server can't load the game in 30s, the client times out
People are connected at the server at the same time
we had a late game save before update 6 and it worked fine too
Upping the server to 16GB got it working - thank you!
thats not possible and a huge security issue
just run it with user 1000
you can still use the root user for docker management
yeah its a bug in the connection.. had the same issue today...
you can't really do much about it besides upgrading and upgrading the cpu...
or storage 😄
Really? I use isolation docker container, really, this is bo@l sh@it, i hope they created special options. This is my choice 🙂
we talked about that on github..
No, this manual about “how to start docker by not root user and link user pid same as user in local” not, i start full isolation docker container, login as root and try start dedicated server as root user, but server decided that i cant use root, this is crazy 😬
its not, if you use isolation within name spaces, root isn't required anyway.
All that i need: option for ignore this stupid error 🙂
I don’t want create more users, root as default
then pull the code from github, change the script, build the image and you are good to go.
when you install docker, it creates a user and a group anyway, so why don't use it.
And take more head ache, sync permissions, sync pids users, really?
I wont use root! No more, no users, no troubles with permissions! I want use in my docker container or virtual machine root user, and I know what i do
if you know what you are doing its no problem to change the scripts and peng you can change everything you want.
otherwise live with what the community has to offer
Face palm…
@hexed gust
https://github.com/parkervcp/eggs/issues/1858
You need to ensure the container is not being started with root privileges in your specialized environment. Unreal Engine servers (which Satisfactory uses) do not allow them to be run with root privileges for security reasons. Nonetheless, this issue is not egg related; the blame can either be placed on your container configuration or how the Satisfactory server is compiled.
so its not even up to anybody on this server, you should ask epic games about it 😄
when you issue the 'stop' console command, and the ds server window is still open, how do you gracefully shutdown the server?
Nah, that's a UE-generic bug. :D
ARK has it, CONAN has it, DnL has it...
Haven't seen that error in a looong time though
Play modded ARK - you'll get reacquainted right quick.
Yeah, I'm staying away from that monstrosity. Tried it but uninstalled pretty quick and that was vanilla
If you have both "Pause when empty" and "Save on player quit" options enabled, you're going to have a bad time when connecting to an empty server. :D
I'm actually quite interested in the server hardware specs of people having that problem. (and the size of saves)
Always wait 30 seconds after leaving the server or timing out. And I recommend disabling the "pause" option - it's a dedicated server, the entire point is that it runs when empty! ;D
Well, one of.
AWS t3.xlarge instance with 30 gig ebs storage 125 mbit throughtput and 3000 iops 😄
I haven't set up that server. Quite sure it does pause when empty. But that doesn't seem to impact my problem. I had the issue when connecting on an empty server just as much as when i have my friend connected
And: I'm running an i7-3770 and I don't have those problems.
What does its single-core speed go up to?
2.5 ghz
but only looking on that as half the problem.. you need to look at the ipc (instructions per cycle), that comes from the general cpu generation...
so 2.5 ghz could still be much much faster then 3.5 ghz on an older cpu
I'm not having problems myself, my CPU is Intel(R) Xeon(R) CPU D-1541 @ 2.10GHz
At any rate, the game has a heavy single-thread bias.
Let me reboot my server see what happens to the CPU during load.
with nvme ssd?
I will ask my friend tomorrow what's the hardware on his server
That technically doesn't matter that much, though I suppose it might while loading.
I'm using 4x 1TB SSD in striped mirror (ZFS), so performance will be quite up there.
I think my problem comes from the storage.. should try a root disk to run satisfactory
Saving is not an issue. I'm saving to a RAID-1 on spinny magnetic wheels.
What are your save sizes?
2 MB? on average, with a max. of about 5 on one session.
754 kb 😄 so really small
Damn, I forgot dism can be such a princess. xD (though on the whole, it's actually fairly simple)
but I think the problem dosn't come from loading the actual save game, more about loading game data as a whole..
will try with gp2 that has smaller iops but more throughput.. with 250mb/s
So I just rebooted my whole container, so the map is NOT loaded. I will join now and see what happens.
If you have "Pause when empty" enabled, it doesn't actually load any game data until you connect for the first time.
In general, I recommend disabling the pause feature.
it is disabled in the compose file
I don't know anything about compose files. I'm talking about the server setting.
How you get there is none of my concern. :)
I do have the pause active and the server took ~20 seconds to load the game files and go from 2.5GB memory to 10GB of memory. It seems that's whats taking the time. It's limited to a single core for me running 100% doing this.
services:
satisfactory-server:
container_name: 'server'
hostname: 'server'
image: 'wolveix/satisfactory-server:latest'
ports:
- '7777:7777/udp'
- '15000:15000/udp'
- '15777:15777/udp'
volumes:
- /dockerdata/satisfactory:/config
environment:
- MAXPLAYERS=1
- PGID=1000
- PUID=1000
- STEAMBETA=false
- CRASHREPORT=false
- AUTOPAUSE=false
- AUTOSAVEINTERVAL=300
- AUTOSAVENUM=10
restart: unless-stopped```
thats the whole compose file
Like I said, heavily single-threaded.
So I guess it all depends on if that process is fast enough if you get disconnects
That tells me nothing.
That may take longer with bigger saves, though.
Which is why I recommend it be disabled.
My current save is around 1MB, but I also measured this a few days ago with an empty save and it was practically the same
Okay. So let's keep looking at those save loading times.
Still, could influence it, but it looks like it just caches all the data in memory regardless of save
Maybe it is the same? Would be interesting.
Though that would mean that people who get reproducible disconnects on load will keep getting them.
The server will keep the memory active when the last player disconnects though, it doesn't unload anything, just pauses
It does unload, actually.
It takes a moment, but it does.
You can see it on subsequent empty connects. It goes through the whole song & dance of recreating the world.
Game files seem to stay in memory though
I don't know. All I know is that you can see it load something, and you get a timeout. Again.
If I reconnect now I instantly get in, server is empty.
Give it 5 minutes, repeat test.
Sure
Paused servers aren't my strong suit, because I don't use the feature at all.
Part of the point for me is that I can come back later to a full container (I do mixed production with lots of manual management).
Electricity is €€ over here, so I'm not keeping the CPU going at 100% when nobody is active 😄
It's actually significantly less when it's empty. xD
Something like 50% at most.
That's of one core.
Well, while pause, it's around 3% for me
Meh. I don't care, I suppose.
Yeah each to their own, it's not like one is better than the other, just preference I guess
Obviously, not having the pause enabled is saving people from some problems as workaround
I see it from the other end: the pause feature is creating issues for people, due to how it's implemented.
maybe loading from a ram disk is the way to go for me.. never done that, but seems to be fun. so 32 Gigs of RAM should be enought to handle that.
Alright, it's been a good few minutes, going to join again.
Ok, so activating the game on the server when I joined took ~2 seconds and then my client connected.
Okay, so I don't know.
I'm pretty sure it's just the initial loading that takes long as it needs to cache all game files to memory. This will cause a timeout for the client if the server performance can't make it within 30 seconds. After the first load, everything stays cached in memory of the server so clients should be able to join without a problem.
It's actually less than 30 seconds, I think. I've seen people time out on 20-second loads. I'm not sure, though, since it works for you.
Yeah I'm still N=1 but the guy we had on the container GitHub issue his video clearly showed the slow loading causing the 30 secs timeout during initial load.
He also confirmed that the second connect, after letting the server finish the loading, worked fine.
when I first got the problem, the server ran 2 hours without any user, I came back and got the time out. So the inital loading was done 3 days ago at that point.
Although I think that we shouldn't exclude the possibility that there are more problems
It would be good to check the memory that's active and see the console when that happens
If you can reproduce it
the console showed me this, active memory was 10.05 gigs at that point. didn't change over that past 2 hours.
when I tried to connect after around 2 minutes, the server tried loading something
So, before you joined it was still at 10.05GB?
yeah, cloudwatch didn't register any change in active memory usage
Interesting
but I think the main problem is with the storage speed and the io bus on the cpu, even while being connected and playing, I saw a cpu que over 1 in 1 minute, 5 minutes and 15 minutes.
so that tells that at least one task i wating for something.
saw something similar with a maria db / apache instance when big querries ran and basicly stalled the cpu
when upgrading from epyc (t3a.xlarge / 2.2 ghz) to intel xeon (t3.xlarge / 2.5 ghz) the problem was gone for 2 more hours and then it came again
Hmm
that just dosn't make any sense to me.. so the only variable to change is the disk speed.
or maybe it depends on the noisy neighbour problem 😮 😄
Yeah CPU shouldn't be a problem unless AWS has overprovisioned CPU scheduling on the hypervisors they run
There is a possibility that it hangs waiting to get a scheduled CPU
But that would be very not nice of AWS
its might not even something the host provider can controll, when the io bus on the cpu is full, due to many database operations (witch is a common thing to run on those small instances) then you wait and wait and wait 😄
we saw that on an IaaS private cloud infrastructure. when changed to dedicated IaaS the problem with slow disk speed was gone..
Yeah makes sense, but unfortunate
I’m running on a 3950x and 64GB 3466 ram kit with no issues at all
And storage speed doesn’t matter for this game btw is only used when saving
And of course initial start up to load the ram
Yeh, that's the main problem with the client timeouts right now, lots of people having that.
Oh yeah, that's a great benchmark. xD "I don't know what y'all are saying - my Porsche Cayenne has no problems reaching 150 mph on that road!"
Speaking of... maybe I'm biased, but this makes me laugh.
Anyone got any ideas on.
If on Petrodactyl
You can use a differnt port than 15777 and 15000. For example use like 25588 for it?
Wed like it to work on just 1 port if possible but have no clue
The Dedicated Server requires three ports you can read how to configure them on the dedicated servers page https://satisfactory.fandom.com/wiki/Dedicated_servers
but im asking can you Change them.
We are trying to host mutliple servers. not just 1
...Did you just not read what I said?
The page goes over how to configure them
Append
-ServerQueryPort=<portnum> -BeaconPort=<portnum> -Port=<portnum> to the launch arguments
Question, did anyone else manage to get more than 4 players working on a Non-dedicated server this patch using the Game.ini workaround detailed in the wiki?
What's the best practice for moving a Satisfactory server to a different physical server?
After update 6, my previous VPS could not support the Satisfactory server anymore and it would crash almost immediately. I tried copying all files over to a different VPS that did meet requirements but after starting the server and trying to connect to it, I receive the below error in my server logs. I am 99% sure it's not firewall related, I can't connect to it even when I disable the server's firewall.
[168]LogServer: Error: Could not open new route from '[::ffff:my.current.public.IP]:53938'.
Question, has anyone been encountering issues hosting a dedicated server on G portal?
Our issue
We just setup an instance and it's kicking out players on join and removing the world from the server (need to restart on GPortal and then recreate world, only to have the same thing happen)
it might be a Gportal question, but just wondering if anyone's seen this before!
Probably RAM. It should save the world even if you time out/get kicked. If it's not saving it, it means it crashes.
This game uses a shite inbuilt router. Use -DisablePacketRouting, -multihome={your server IP} or -multihome=0.0.0.0 to get rid of it.
The latter two, as long as they're provided an IPv4 address, will also kill all IPv6 this server gets confused by. xD
I mean, what the fuck? Routing an IPv4 incoming connection via IPv6, to reach... IPv4? xD
Dude, thanks! multihome worked. But the server now shows as unclaimed, it won't overwrite my previous save data if I reclaim the server?
Lol hahaha
So reclaiming the server does not seem to find any of my previous save files. I guess I'm not even sure where exactly those are saved server side.
My old server can no longer run a Satisfactory server. I do have access to all server side files though, I thought copying everything to a new server would bring save files but I'm only able to start the server and create a new game.
It shouldn't touch any saves, no.
They're in the user's home directory. It should see those saves - assuming you copied them. Copying the server directory is not enough. Under Linux, they're in /home/{user}/.config/Epic. Under Windows, that's somewhere in %APPDATA%.
in windows is in \AppData\Local\FactoryGame\Saved\SaveGames\server
Dude..... There's a reason why I named this save file the way I did... Thank you! This only makes me ask more questions that I don't want to understand why but I think I have what I need now.
@glacial glen
Is there a way to not have new players entering a mature server start from scratch for themselves?
Explain?
If they need stuff to build, they can just grab whatever's available from any containers, right?
Yeah. I mean if someone on the server has say unlocked everything, then a new person joining does not have the urge to play to unlock as it’s already available.
My question was that is there any way in config to have this reset for new players
Ah. You said "not start from scratch".
No. All progress is shared.
Everyone is essentially on the same team.
Ahh OK. Bit of a shame. But I see the working of the system now
I guess what I was looking for was that every player has their own hub
Not an option, sorry.
I wish it were possible as well, sometimes.
Still, having separate "tribes" would come with everything that entails -- being able to block someone else, allowing squatting in an area/on resources (think: someone quickly hops around and puts miners on all uranium nodes, or all sulfur nodes), or just losing access to your shit if you happen to spawn with a new ID (which does happen, from time to time). Coupled with the missing console commands, there'd be no way for you to get back in - you'd have to start a fresh save.
And yes - you could evict the asshole who did that (by changing the password - no other access control exists), but you'd also have to start a fresh save.
This sort of division also doesn't work in lore: if you're all FICSIT employees (more like slaves, it looks like...), why would you compete against each other, to the detriment of FICSIT?
As it stands, everyone is supposed to work together and have a gay old time raping the planet and polluting it without end. xD
Actually... if I didn't think FICSIT would take everything worth anything - including the water, the remaining grass, the sand, and the atmosphere if they could - I'd say they'd just leave everything as-is when they split, without half a fuck given to how much the remaining machines will destroy the ecosystem until they finally break down.
But that's sorta neither here nor there... :D
Is there a way to disable friendly fire?)
As far as I'm aware, no.
ben I did that. Doesnt work lmao. Thats why im asking xD.
To see if im missing something.
If i set all 3 to the same port it fails to work.
the port i have set is 25590 not 92
./Engine/Binaries/Linux/UE4Server-Linux-Shipping FactoryGame ?listen -Port=25590 -ServerQueryPort=25590 -BeaconPort=25590 -multihome=0.0.0.0 $(if false; then echo "-DisableSeasonalEvents"; fi)
is that even suposse to be working? traffic mix up and all that good stuff
how do you tell the client to use different ports? 😬 is that part of the first connection to the beacon port?
hey i bought a serveur but cant find him in game
this is a very common problem, even with the best config it is still a problem by the server Data. we have this problem to as the Experimental version starts
where is the saved games stored for steam now? it use to be under app data but not anymore. anyone?
Since the ports are all UDP, you can't have them the same - each has to be its own port. The server detected this conflict, and automatically increased the ports until they fell to 25590, 25591 and 25592.
The server tells you. As to how, I have no idea. Query sends the Beacon port, if memory serves, but other than that it looks like encrypted traffic (which it probably is).
here are the server specs :
https://cdn.discordapp.com/attachments/962820005546831922/1023883982733053993/Capture.PNG
yeah the same as my server 😄
what about the storage?
It's a virtual machine but it's running like 10 cores + 16 gigs of ram I am told
Yeah but cores don’t matter for satisfactory DS is all about single core performance
It seems most people that are having issues from what I’ve seen are using very old xeons from like 2012 or so which are very low in terms of single thread performance and IPC compare to newer cpu
To be fair, those old Xeons still kick butt quite soundly. It's just that they get to shine in multithreaded applications, now that most of our stuff actually is properly multithreaded. You could say they're seeing their second golden period.
But yeah... about as good in single-thread-biased stuff as they ever were, I'm afraid.
I wouldn't say "no" to owning a sever with dual Xeons in it, for development purposes. Game servers however... run about as well as they do on my old FX-8350.
@blazing halo get a i9 12 gen
The FX-8350 has been retired about a year ago in favour of an i7-3770.
I'm not about to drop dough on this, at any rate - the only reason I'm even pushing for upgrades at all is because my father wants to play multiplayer games like ARK.
For the stuff I need to run? An armv7 would likely be more than enough.
It used to be a Phenom-II X4 before that (while the 8350 was my main rig), and even that was more than enough power for my purposes.
Thanks, but I'll pass. The i7 is more than powerful enough to manage.
I also try to not run on someone else's hardware. Unless we enter into a formal agreement, I'd rather not - I'm not chomping at the bit for disputes and drama that inevitably arise. :D
Aww I understand sorry 😞
If you were willing to run 19 separate instances in a cluster for free... 😂
... I might reconsider. 🤣
You'd need some 256 GiB of RAM for that, though.
I have 196 lol
Just for that, mind. Each instance is conservatively about 14 GiB.
Yeah, well, I ran 'em two at a time whenever we wanted to transfer servers, so it's not like I had trouble with it (though 32 GiB is cutting it close). Which is why if I were to move, I'd want an improvement.
Also, aren't you somewhat halfway around the world from me?
(if you mentioned, I already forgot)
I dint mind hosting for people as long as it makes them happy
@blazing halo server is in Germany
Ah. That's very close, actually.
Still, I'm reluctant, especially now that we're not playing ARK as much.
I understand offer always stands
I keep it in mind, as you may've noticed.
I did name-drop you once or twice when hosting was discussed.
I hope that's okay, by the way?
Yes just busy almost all the time lol sorry 😞
I ram 12 clusters on 64GB but it was pretty much maxed out but the cool thing about ark if they are is no players connected ram usage goes down on the clusters
You might want to draft a basic service agreement and price it at, say, $1/mo for a maximum one-time allocation of 6 months (extendable in perpetuity). I find it much easier to deal when I know there's a formal agreement and a provider/customer relationship in place.
Most importantly, I don't like guessing if I should ask for something (or even demand it) when it's a "gentleman's agreement", you know?
I prefer it when it's spelled out and written down.
So, while we may be friendly, I'd still prefer if we could treat it as a professional contract, you know? And I think there might be others like me.
Food for thought.
I mean, nothing's stopping you from doing it entirely pro bono on a "gentleman's agreement". Just... might want to consider having the contract option.
If you're serious about having people use it, I mean.
And please don't take this the wrong way - I'm not demanding anything of you, just suggesting. I've been burned a few times, on both ends of this issue (I used to host websites 15 years ago), and I'm now much more cautious about situations where the responsibilities, rights, privileges etc. aren't clearly defined on both sides in ways that make it easy to reference.
Hi, I hope someone can help me, I have my own dedicated server runnings Win Server 2019 Std, I have followed the instructions on both here and the Wiki but am unable to connect to my Satisfactory dedi server.
It seems the game clients makes a connection where I can create a new game but then when connecting to the server to start playing, it then fails with this error message:
You might need to post the server log for us to help you more… just make sure you have more than 10gb free to be able to load the map.. minimum requirements are 12gb of ram now for update 6
I see this in the console on the server after that message comes up on the client
The server has over 300gb free
Thank you for replying so quickly though 😄 appreciated
We get the same error message now. Worked fine for 2 days. Renting server from indifferent broccoli.
I don’t mean disk space I was talking about dedicated RAM
Just to verify as most people don't have more than 128GB RAM in their server, you are actually stating RAM/memory and not disk space right? @pearl pagoda
But just post the server log files is a small .txt file you can just post it here
should be good 😉
Oko, can you direct me to the log file please?
these ones?
full disclosure, I have Empyrion, Rust, 7D2D, Conan & Valheim servers all running on this same box, I can connect to them all no problem
That seems like your having a networking issue to me
From what it seems the server started fine
Is when your trying to connect something with your networking is not allowing it
indeed, as i say, it makes the initial connection fine, i can setup the game, its only when joining that it fails
Well the default timeout is 30 secs if the server doesn’t get a response within that time frame it will give you the timeout error
Maybe increasing that timeout could help but hard to know
How do I increase the timeout?
Thanks, it seems to be out of date though, I cant find any of the files its refering to
If they aren't there you can create them. It's Unreal Engine configs so it will look for them at boot
😦 no change
I set the timeout to 60s but it only took it 10 to fail to connect and throw the same error so i think its something else
Pick one of -DisablePacketRouting, -multihome=0.0.0.0 or -multihome={ip of the server} and apply to the server's launch arguments.
-multihome=0.0.0.0 worked!! thank you @blazing halo 👍 you're a genius!
No, but I play one in front of a camera. ;]
I was about to give up on this so, you've made my day, thanks again!
(which is actually easy, if tedious, to do - you can do it over as many times as you like, after all)
To be honest, my satisfaction for sniping the issue is my reward.
But no, it's just experience.
Besides, others have eliminated a few other possible causes while I wasn't looking, so...
yep, first time running a Satisfactory server, I can get an Empyrion server working with just some sticky tape and a bottle cap, as you say, its all experience.
So this is the first thing to check right… good to know future reference I was to link him this but just little busy at work lol
Basic rules so far:
- having trouble connecting when you by all rights should be able to?
disable the internal packet router first, see if issue repeats - if you're using the "Pause while empty" option, expect a timeout on first connect
not avoidable, sadly - if you're using the "Save when player leaves" option, expect to be timed out if you try to rejoin as soon as you can (while it's still saving)
wait 30 seconds before trying again - the server is highly biased towards single-thread performance
increasing tickrate won't make it go any faster, but it WILL make it lag more :D - 12+ GB as of Update 6 EA
- NO TUNNELS OR PROXIES
no exceptions - the game needs 3 ports; they're all configurable, and they're all UDP
don't open thousands of ports to effectively DMZ your server - just don't - remember that Windows comes with a built-in packet filter (some call it a "firewall") that is enabled by default
thing is, I think I read this in the wiki but i read it as optional and only necessary if your server has more than one interface, as mine only has a single interface i figured I didnt need to specify it in the launch arguments
If you see trouble connecting where they by all rights should be able to connect, and you see ::ffff:127.0.0.1 or some other IP confusion shite - assume the internal router is at fault.
nice, good to know! thank you again
Yeah, except if you have IPv6 enabled, it may confuse itself into using it, and then you get shite like a connection to 10.0.1.22 going through ::ffff:127.0.0.1. I'm sure you can see the absurdity of that?
indeed, well its new and shiny, hopefully it will get more .... "robust" in future versions 🤣
Thing is - the game has a built-in router. You connect to one port (Query, I believe?), and it routes the packets among the other two.
Except when it doesn't work.
And it often doesn't.
Hey, in the latest update we were granted the ability to change the Beacon port. I'm ecstatic.
lol, its the small things
They still silently auto-increment though, so watch out for that.
I don't have a tendency to open 1000 ports if the game uses only 3, so if you're like me - it might auto-increment beyond the forward, and you're cooked. ;P
Anyone got any clues here.
Currently using ports 25589 25590 25591
The Create game comes up. But once i start one. It just breaks. disconnects me from it and i re-create it again. Its a loop.
When i log in it does the Ficsit tips. But doesnt get me into the game. and runs outa ram.
We are even trying the default ports and It doesnt work..
Pterodactyl panel is what we are using so it should work. but the egg i guess is messed up.
I'm running on Pterodactyl, "runs outa ram", how much RAM did you assign?
we gave it 12gb
That should be enough, do you see it hit 12GB? It should stop right about 10GB after the first player connects or on map load
The one from parker.
My dms are open if needed lol
Can you get me your startup command?
Mine:
./Engine/Binaries/Linux/UE4Server-Linux-Shipping FactoryGame ?listen -Port=7777 -ServerQueryPort=15777 -BeaconPort=15000 -multihome=0.0.0.0
Variables
I connect to server <ip>:<query port> on the server manager
Thats what im doing
The issue is halfway though it disconnects and says connecting. aOR errors
also make sure your allocation (under network) has your "game" port as primary (mine is 7777)
Waity might of done it
We are tryna make it so we can host for more than 1 person lol
Yeah
You can set game port to whatever
mine just happens to be 7777, which is the "-port" option
all 3 ports need to be separate, and -multihome may or may not be needed depending on your setup (I just haven't touched mine)
there is a way (or was) to use just the beacon (or query?) port to connect but I never had luck with it personally
You're supposed to connect to the QueryPort, connecting to other ports should not work.
Interesting you say you connect to the Gameport? 😄
@vale parrot I'm talking about the internal routing thing they did via the query port.
As for "game port" in pterodacty designate as primary, and then the parkervcp egg assigns that to the -Port=X property
Yes, if I connect to that it fails
I never said to connect to game port 😛
Yeah, was (maybe is) and it was the query port, for some reason I thought maybe beacon port possibly but I never got the routing thingy to work myself (using server query port). But I also haven't tried in a long time, and I only host 3 or 4 game servers right now, so I wasn't worried about port numbers.
There is, yes - you turn it off with -multihome=...
Please elaborate on the "and runs outa ram" bit.
And yes, that does track with no world being saved - it has to save it first before it'll be able to load it later. If it never makes it that far (because of, say, an OOM crash while loading), you have to recreate the session from scratch.
Does anyone here know how to update a containerized server from Update 5 to Update 6? Specifically this tag: ich777/steamcmd:satisfactory
Evening all, quick question as I beleive Ive done everything correctly setting up my dedicated server. Do other players need to open up their 15777 port as well to join?
no
Hello, I'm server admin. How can I cheat this game?
give items or spawn , Can I do anything?
hi everyone, i just looking for some hosting for me and my firends, any sugestion for easy setup hosting in singapore?
Rule-break?
Usually you dont update containers, you kill the old one and start a new container. The new one will pull the latest image before creation and should install the new version. But that only happens if everything is set up correctly so I would doublecheck the image and see if your persistent data is configured before you lose your saves on the server.
Nope there is no admin command like that
how do i create a dedicated server? I want to start fresh and yes i've looked at the wiki
Did you only look or did you actually read it? It tells you everything you need to know 👍
pls help what do I do?
Launch the mod manager and update that and your mods first
Anyone had issues with attacking spitters and the game suddenly becoming a slide show on DS? Frame rate drops significantly. Haven’t had any issues with multiple hogs/spiders but both large and small spitters are almost unplayable. I searched and will post to QA but wondered if anyone was having the issue?
It seems to be a bug when interacting with projectiles from spitters, it's not really a slideshow, just extreme rubberbanding.
Ahhh gotcha, I assume we just post it to QA? No workaround atm?
Yep, not anything I'm aware of, just dont get hit haha
Is there any issues if I tried to connect from a epic games account to a steam dedicated server?
not really i do it practically every day
that i have no idea i use steam cmd
Hi all, tech noob here:
So I created a dedicated server (same machine running server and playing game), but I can't get people to join
Also when I search the ports I've opened (arris router) on canyouseeme, they don't show up
Does anyone know what I'm doing wrong
I've set static ip, ran the port forwarding (all 3) on the same static ip, allowed thr query port (15777) on the firewall
Can you connect on the same LAN?
Tried it with radmin vpn and it didn't work
Can be a multitude of things, did you create rules for windows firewall?
I created 2 rules: one for TCP 15777, and one for UDP 15777
I should mention that port 7777 is being used by something else (I think media service center)
And is it at all listening on the port? you can check with cmd: netstat -abn | findstr "15777"
Gonna try that
When I ran windows server I remember that I had to use: -log -unattended -multihome=10.0.30.10 as parameters. Ofcourse use your own ip ^^
Not sure if i'm helpfull at all, I was looking for what 7777 did and it's the game port which you could redirect using the -port parameter, all explained in this link , which is also a good troubleshooting starting point: https://satisfactory.fandom.com/wiki/Dedicated_servers#Configuration_issues
is it normal for the game to keep freezing in a dedi server
Not sure what's up here, but I've got a Linux dedicated server instance that's been running for a while. (though not really in use a lot before U6) I've been off and on it since U6 and I've noticed a couple times the server boots me. With this last boot I was on a train. Now the server via command line reports it's up, but Satisfactory reads it as offline.
Seems to be using wuite a bit of memory and CPU actively. Here's output from systemctl.
`hiro@stormwind:~/.config/Epic/FactoryGame/Saved/SaveGames/server$ sudo systemctl status satisfactory
● satisfactory.service - Satisfactory dedicated server
Loaded: loaded (/etc/systemd/system/satisfactory.service; enabled; vendor preset: enabled)
Active: active (running) since Tue 2022-09-27 12:06:55 EDT; 8min ago
Process: 1602 ExecStartPre=/usr/games/steamcmd +login anonymous +force_install_dir /home/hiro/.steam/steamcmd/SatisfactoryDedicatedServer +app_update 1690800 -beta public validate +quit (code=exited, status=0/SUCCESS)
Main PID: 7175 (FactoryServer.s)
Tasks: 32 (limit: 18955)
Memory: 13.0G
CPU: 11min 29.934s
CGroup: /system.slice/satisfactory.service
├─7175 /bin/sh /home/hiro/.steam/steamcmd/SatisfactoryDedicatedServer/FactoryServer.sh
└─7183 /home/hiro/.steam/steamcmd/SatisfactoryDedicatedServer/Engine/Binaries/Linux/UE4Server-Linux-Shipping FactoryGame
...
/Game/FactoryGame/Map/GameLevel01/Tile_X5_Y1.Tile_X5_Y1:PersistentLevel.InstancedFoliageActor_0.FoliageInstancedStaticMeshComponent_78)
Sep 27 12:07:57 stormwind.slasho.me FactoryServer.sh[7183]: [2022.09.27-16.07.57:052][ 1]LogNavigationDirtyArea: Warning: Skipping dirty area creation because of empty bounds (object: FoliageInstancedStaticMeshComponent /Game/FactoryGame/Map/GameLevel01/Tile_X5_Y1.Tile_X5_Y1:PersistentLevel.InstancedFoliageActor_0.FoliageInstancedStaticMeshComponent_94)
Sep 27 12:07:57 stormwind.slasho.me FactoryServer.sh[7183]: [2022.09.27-16.07.57:052][ 1]LogNavigationDirtyArea: Warning: Skipping dirty area creation because of empty bounds (object: FoliageInstancedStaticMeshComponent /Game/FactoryGame/Map/GameLevel0
UPDATE: Ignore the above. Restarted game client and it connects now.
i got lost in this wiki trying to figure out how to check the ports lol
i think i tried netstat -na
and was seeing the ports (but they did not indicate any status(i.e. listening established))
If you see them then the satisfactory server is setup correct to receive incoming traffic
If you have a second PC you could try and connect and see if it works in your LAN.
If that works then it's probably only the portforwarding that needs attention
Tbf you mentioned that canyouseemee is not working, and netstat is showing you that the ports are open. Either windows firewall or portforwarding I suppose.
Could it be an ISP problem?
You can't.
It's not a slideshow - you still get however many FPS you should. It's a collision/world-object movement issue.
If the last thing you caught it at doing with journalctl was LogNavigationDirtyArea: Warning: Skipping dirty area creation because of empty bounds (object: FoliageInstancedStaticMeshComponent /Game/FactoryGame/Map/GameLevel0, then it was still loading the save/world. Restarting the client probably didn't do anything, except delay your connection attempts until the server was ready - and then it worked.
A bit of networking tools knowledge would be handy to have. nestat (and ss and so on) only show ports that are being actively used. If they're not used, they're not shown.
Though you may want to limit it a bit with -l I think... (and further with -u if you can (and I'd recommend -p to see the process identification...)).
tl;dr git gud
ZanGi is trying to "git gud", we al have to start somewhere
Do you even know if you have a WAN IP that is publicly routable? ;)
Bruh
Yeah, yeah, yeaaaaaah, I know.
We all start somewhere. Of course we do.
So I'm assuming maybe I just might not have
So how much networking know-how do you have?
I have basic coding knowledge
You thought maybe STARTING SMALL? XD
And that's the extent
Buy a server? xD
The frustration of making a dedicated makes me want to buy one
But the reward though
To be fair, I don't have anything against people who don't have the knowledge. It's just... "How hard could it be, amirite?" :D
I honestly don't think it's hard it's just that when I encounter a hurdle it's a bit challenging
Yup
Guess it makes sense when you put it in perspective.
"How expensive could it be? I just gotta buy the car, and pay for gas sometimes, yeah?"
If you've had more than 1, or at least for a while, you know exactly how wrong that statement is. xD
But eh.
I'm in the auto business so I know all too well lol
I've been told that this channel is mainly for helping people, not discussing philosophy. Sadly. xD
So.
Please go into your router, and locate your WAN IP.
But going back to "you don't know how much you don't know": idk where to start troubleshooting
(if you don't know what "WAN" or "IP" means, Google it - I expect you to work for it a bit)
Note your WAN IP down.
Next, go to a website that tells you what IP it sees when you connect to it - whatismyip.com works OK for this purpose.
Note that value down as well.
If these two match each other, you have a public IP.
If they do not, you will need to talk to your ISP and ask them to assign you a public IP. They may demand a rationale.
a you're going for the [pc] --> [router] --> [modem] where the hell are you at approach
A "public IP" is not a thing that you have. An IP can have the property of being "public".
Ranges 10/8, 192.168/16, and 172.16/12 are specifically private and cannot be used on the public Internet.
and vise versa, dont use public ranges private because they look cool xD
Some ranges are also reserved, but you likely won't run into them. The exception is 169.254/16, which is the zeroconf/autoconf IP.
Had a guy once with a 15.x.x.x subnet apparently that's the range for the public HP websites, wondered why he couldn't go to their website
Hamachi used to use 5/8. I still don't know how they got away with it.
pure luck
@cedar crystal and?
any idea why its so unplayable
Looks like a client problem. Does it happen in single-player too?
Then please test in single-player.
Looks more like a CPU/RAM issue. Or maybe disk, if you're running the game off of a spinny metal-coated-glass surface.
i think the game is installed on an external usb hard drive ngl
only thing that had space
ig ill install on the sd
ssd
Installing games on external disks is usually not a very bright idea. xD
Speaking of not very bright ideas,
anyone remember DVD-RAM? 🤣
im reinstalling game now
@blazing halo so the WAN ip in the router and whatsmyip are the same
Congratulations, you seem to be a proud owner of a public IP.
Okay. So now explain to me why you allowed just the one port on the firewall, when you knew to forward all three on the router. ;]
Also, since the server now allows us to freely configure the three ports, I would recommend making them sequential - for example: 15000, 15001, 15002.
I suppose I may not be able to blame you, if you read somewhere that you need just the one... the game has a built-in router, true - but it's baah-rooken and worth about as much as my old Netgear one. The one I turned into an ashtray. That one.
(you can disable it, and the switch that usually works is -multihome=0.0.0.0 (hope you have no love for IPv6))
I was following a tutorial on YouTube where he forwarded: queryport (15777), beaconport (15000), game port (7777), and then he just made the rule for that one particular 15777
You're probably aware of those tutorials as well lol
So I guess I'm going to change the port forwarding to 15000, 15001 and 15002,
Then in firewall set rules to allow each port (UDP & TCP)
UDP is fine, there is no need for TCP
Sure makes it easier
I wasn't, primarily because there's an excellent Wiki writeup in this channel's topic.
@cedar crystal remember to actually configure the server program to use those ports
is there a update 6 dedicated server for a docker?
Another thing: should I have steamcmd in the same folder as Steam?
hey folks, the dedis wiki page says
To download the Experimental branch of the Dedicated Server, add -beta experimental to the end of your SteamCMD invocation. It is not yet possible with Epic Games Launcher.
however,
is this new?
Looks like just how EGS does it?
uhhhhhhhhhhh
so i logged out while in an explorer
and then the game updated, and now im under the map
Shouldn’t have saved while in the explorer XD seems to not be an issue in SP, but I guess it is on MP and dedicated.
yes
Still cant see drones or radar towers on my map.. has anyone else had this problem.. if i switch to single player i see them no problem.. but in the server they are not showing, and neither is the options to show them in the map menue.
i did it!. i dont know if there is a guide anywhere for it but i figured out how to choose file location, auto install, verify install, close steamcmd, and then launch satisfactory server, all in one click. if there is a guide out there, i would like to see it if there is a way to clean up my command list
Congrats, guess you are on Windows?
You can always try to see how containers are set up, they have everything like that in an entrypoint or start script, might help you to get some extra idea's.
This was my first update using a dedicated server. How do I update it?
My client is the new version but server is the previous one
hey i cant connect to a dedicated server, i tried pinging the ip at it doesnt respond, any idea on a fix?
I just stop the server and re-start it via steamcmd and it auto updates
Is there a console command or env variable that enables a server to keep running when no one is online?
how do i update a server
How is it deployed? If it's through SteamCMD, you just run the install again and it updates it
its done through pterodactyl if anyone knows what that is
which uses steamcmd so i can probs get it to work
ah yup just restarting it works
does anyone know if there is a way to poll steam for whatever the current version is so i can compare it to mine. i have a command set up so it will auto exit the program and relaunch to update. or does the -autoupdate command for the steam batch actually work
You can query the STEAM API to see what's the latest bundle change, sure.
Is there actually a server there?
The same way you installed it.
I wasn't aware that steamcmd is now in the business of managing system service states 🤣
General Computer Knowledge 102: never keep two independent programs that use the same files, folders, structures and/or resources in close proximity to each other, unless you know exactly what you're doing. That's just asking for one to mess stuff up for the other. It can be beneficial (it may help you save on disk space, for instance, or on updating time if one of them can do updates that both can use), but that's a call you can make when you have intimate knowledge on how they both work - and even then, more elaborate setups (separate locations with soft-/hardlinks, filesystem mapping of specific files etc.) are often safer... though admittedly much more involved.
yes
Where's that host located? (network-wise; I don't care about its geographical location)
do you have an example for the api poll command? i dont know where its going
[h]ttps://partner.steamgames.com/doc/webapi_overview
holy damn this is new territory for me. if this works i will clean up the batch and show you the file
You'd need to query the storage API to get the change number, and not every change number indicates new files to download IIRC.
You can also try parsing [h]ttps://steamdb.info/sub/17906/ (that's the public server package)
Not sure if you need to query the specific branch too, since this game has betas.
im looking at running a curl command for a specific line on the api site
You're still going to get JSON in your reply, so you'll need to parse it anyway.
then comparing that number to the current number listed on the existing file. if they dont match it should close the program, and launch the batch i have to update and relaunch
thats fairly easy
Yeah, except that's not a number you have in any package, IIRC. Maybe inside STEAM stuff.
The game build isn't listed.
For example, the last changeset number is 16185040
i can make a text file that has it in there and use it as a trigger point
then when it updates the game, have it update the number in the text file
You can. As long as you keep it up-to-date, you should be good.
Of course, if your updater terminates for whatever reason, you're cooked.
So I'd recommend keeping some more state. :)
look, screw precaution, live or die lol
Not my style. :)
im learning all of this on the fly so in fairness, i dont truly know what im doing
That much is obvious. :D
I would recommend using some programming or scripting language.
I personally use PHP for stuff like this.
i havent really learned anything more than screwing with .bat files
Well, no time like the present.
i really appreciate the positive response for learning this. most people i come across have been rather dismissive or even hostile at times.
I can be hostile at times, especially when people have no fucking clue what they're getting into. xDD
But eh.
i took this on as a homelab project for practice in a sysadmin position, and my friend thats a supervisor gives me tasks to do that i have to figure out. its helping me alot better than college did for me. (but not my spelling)
That's not actually a bad exercise. Congrats for having someone lucid at the helm. :D
thanks man. tonights project for the automating the whole system has been a huge headache. first time curling
I can probably help you out more later. DM me.
Can you have a dedicated server with update 6 on it. With mods installed. Or is that nog working
Some people are trying to get mods to work, but AFAIK its not possible yet.
Sad
We're still waiting on the mod community to make it possible, yeah.
I dont know if i am right. But i saw somewhere that zap hosting had it working
They don't, and if they say they do - they're either lying or are misinformed.
As for Nitrado, we know they're lying through their teet. They know perfectly well that mods don't work, and they heavily imply that they support them on their offer page anyway.
Ah, I see. "and offers full mod support thanks to FTP access"
Yeah, no.
I fired an e-mail their way and asked them to clarify this position on their offer page.
Having FTP access is great - though from what I'm seeing, the mod community wants to provide tooling as well, and that'll be CLI, so SSH access or some other way of sending commands will also be required.
I wonder if they offer save downloading, because this seems to be a bit of an issue with some hosts.
(it seems that not everyone got the memo that the saves get put in the home directory of the user, and not in the server install directory)
Damn thx for the info
WHY MUST SERVERS ALWAYS BE SO STUPID
i just wanna play
but i needed to update
it says permission denied but if i use sudo it says dont run with root previliges
There is a fine line to updates and glitches isn't there
Big shout to @blazing halo for being such a pivotal resource in the DS discord. You’re a hero in here my friend. I’m glad you’re part of this community and hope you continue to get enjoyment from helping people.
Nice to be appreciated. :)
Setting the right permissions is part and parcel of sysadmin duty.
But yeah, I notice that programmers have problems with setting, demanding, or respecting permissions in general. ;]
In general, it's a good idea to have each such service run as nobody if possible; if not, creating a user is the next best choice.
I know this may seem alien if you're more used to user-side programming, but that's how it is in servers. You'll notice: if all of the servers are "so stupid", maybe it's not that servers are stupid, but you're looking at it from the wrong perspective?
If something is repeatedly done in a specific way, maybe there's a reason for it?
And while the implementation exists to do your bidding - as it always does - it has its own rules and rituals it's governed by.
It is always a sign of respect when a newcomer strives to learn those rules and rituals.
Learning to ask the right question is always a good start. :)
It's also why I think that learning to sysadmin/netadmin is a good thing for a programmer to do; not only does it reinforce certain aspects of programming, but it also gives perspective for the "other side" of things.
i found the error
i did
chown -R Steam *
and reinstalled the server before
Or, you could've just chown steam ServerInstallDir, and then su steam and run everything else as steam.
Have steam create all of the directories, files and so on, as itself.
I would recommend installing screen, unless you wish to run the server as a service.
Also, there's a fork of PuTTy called KiTTy, which can remember passwords.
(I don't recall, but PuTTy can probably remember keys, so if you're using those, then nvm)
Because save big ;P
Also, that looks more like 10, not 13.
Hi, can someone help me ? tryin to create a game on my server, after the loading screen server goes Offline and this appears
What does the server log say? (I fully expect this to be a case of "out of memory"; the server needs 8+ GiB to run)
Hello folks, I am running a server on G-Portal and seems like there was an update to game today/this morning (from .70 to .77) at least I got a mismatch when trying to login. I rebooted my Epic client and my server and able to see the server fine when in the server manage option within game. But when I try to connect the game loads forever and never gets me in game. It briefly flahses a super low rendered screen with graphics, and then back to the black/orange load bar screen and just stuck there. Anyone got thoughts/ideas on what to do, or what might be causing it?
What does the server log say?
I had the same issue yesterday
And what did your server log say?
I am not home currently soo ill check when I'm home
While I'm at it: how much RAM does the server machine have, and how much is made available to the server process?
6gb i
Ooof.
My host said that in order to play it u need to upgrade to 6gb I did and my ram started maxing out 😑
Yeah, Update 6 needs at least 8 GiB to even start a new game, and you should expect to reach 10 GiB (possibly more, depending on how much you build) down the line.
Ahhhhh my wallet
Is going to hurt
Ask the other players to chip in. ;)
That is true
Be glad you didn't participate in U6 beta - that needed 10.5 GiB just to wake up.
Hi, some short Question. It feels like every third patch or so trains are working verry smooth on Server Games. any other the Trains are Stuttering arround and overround them self. At last the Patch before the last one here the trains worked fine.
Are there some differenzes between the Patches that can cause this?
You'll need to ask the developers that one. ;P
(there's no way for us to know, except what we can see)
Did the old college try reboot of my local machine and tried to log back in. Seemed to have done the trick (also server is 8 8GB only currently). I was logged out in a truck, so not sure if that was causing issues. It does have me placed under the world now in the "vehiocles are stuck" location. So there is that
8 GB is not enough RAM. Sorry.
No developers here ? 🙂
No. All of this is mostly community-ran. There are devs on the server, but they're too busy working on the game to answer questions. ;P
However, if you have issues with specific spots on the train rail (always the same ones),
remove and replace that piece of track. Can help.
Got it, ill get my buddy to chip in as well for it. Appreciate the help my friend. Still pretty new to the dedicated server thing.
To be fair, the server's RAM requirement is growing, and will probably continue to. Update 5 needed 6~8 GB.
no its realy Patch dependend not track dependend. I've observed this a long time now 🙂
Then compile your observations (videos and savefiles would probably be very useful here) and post on the QA site.
Is there a way for me to recover the truck/items within it, or is this more of a lost cause thing now. along with any items im carrying in inventory ?
If you died, you should have a death chest that should've been placed topside by the server.
As for your truck... did it fall wayyyy below the map, or is it sitting on something?