📡│hosting-providers-chat

[junior sparrow]

It seems like the game's servers for connections are having issues. We're seeing a few servers through us having connection issues all of a sudden.

[narrow plaza]

We have seen many issues with connecting today. Using direct connect with IP/restarting the server seems to help in some cases.

[junior sparrow]

I tried a bunch. Direct connect didn't work for me either. I ended up tcpdumping the flow and their TURN servers are severing the connection for whatever reason.

[narrow plaza]

Yeah, we had issues with servers "freezing" ie network activity stops and kick out all players as well.

[junior sparrow]

The full handshake shows up, I see the timing logs appear, and then for whatever reason it resets the connection which causes the server to panic and dump a stacktrace lol

[prime badger]

i see that on mine as well

Account already disconnected. State Broken [D:\Source\Build\work\gameRepoCheckoutDir\Plugins\R5BusinessLogicCore\Source\R5Net\Cpp\Cm\GsStream\R5NetCmClient_GsStream_P2pClient.cpp:101]

just keep getting connection state broken every time lol

[narrow plaza]

Have you tried using direct with IP instead of invite code? Saw that help some people.

[prime badger]

yes, so this is a private dedicated hosted from my Dev/Gaming PC that i will be switching to my new Server rack soon. but i can't even connect from my own pc to my dedicated on the same drive.

the loop just fails call and says broken then kicks

i will try direct ip again. i just restarted.

[prime badger]

i used the experimental forced relay and it let me in

without direct ip

[wind tulip]

@narrow plaza When you mention networking issue, do you mean my network? or to the game ? I can tell you it's not my internet as i got pretty beast connection so it's deff not on my end for the servers crashing

[narrow plaza]

Not your network. Something on the game-side seems to be closing the connection.

[wind tulip]

From what i know about how Unreal engine works as i've use to develop on it myself.. The logs provided on every crash seems to be an issue the server trying to load the actors/prefab assets which the game uses be put into the server / game client. Unreal engine uses the language to use the game its server to load these and that connection being cut makes an overload in attempts to load the needed actor/prefab asset.

Not sure if that knowledge has any type of use.. But that's what my experience knows

[narrow plaza]

Let me @smoky sentinel for when hes around. Might be useful information for him to have.

[wind tulip]

Great , For now i'll be off and around tomorrow through the day if needed. If needed i can message my email on DM for which account i got on bisect. Not sure if you guys can get into my server logs to see the crash results but hit me up if you do need them.

[narrow plaza]

Feel free to DM me a link to your panel.

[narrow plaza]

You've been quiet today. Are you guys running into any of these connection/freeze/timeout issues?

[viral compass]

Apologies, we have actually had 0 issues 😅 taken full advantage of a 0 tickets day to enjoy some of the game myself

[junior sparrow]

Do y'all run through Wine or native Windows?

[narrow plaza]

Enjoy! Its fun for sure.

We are not on native windows. Kringle tried a few different things earlier to try and get it working on our end.

[viral compass]

We aren't native windows either, use Wine to run ours

[narrow plaza]

Mind sharing what version for comparison here:#1500810776150278258 message

[viral compass]

Yeah we are running Wine 11.0 :S All our guys restarted for the update as well

Just launched a fresh server and its installed and booted up fine. Testing log in now

[swift pendant]

We are also using wine11 on our open source container, no freezing issues here that I know of

[junior sparrow]

I've not heard of any freezing on our stuff, just the login issue currently.

[narrow plaza]

By "Freezing" its the server closing connection to users and then becoming unjoinable.

stays online tho

[junior sparrow]

We're running Debian 13 containers with wine 11.X whatever latest devel is atm

Our only winetricks are vcrun2022 d3dcompiler_47 and that's mostly there for UE4SS.

[viral compass]

Yep just logged into our freshly fired server fine, no issues at all

[junior sparrow]

🤔 lemme try a fresh server. I've only been trying client servers.

[swift pendant]

Hmm
I've maybe seen a couple of those, most of which we advise them to use direct connect or schedule 1-2 restarts - but no widespread issues

[narrow plaza]

So none of you guys are getting any reports of all players getting kicked from the server after random periods of time? (Minutes to hours in the reports I've been given.) The logs seem to just show the connection being killed, typically after 5-10 minutes.

[junior sparrow]

I'll ask our CSR team if they've seen an increase in it

[narrow plaza]

Appreciate it.

It seems to have started on our end at the same time as everyone else. There are dozens of posts in the discord from various users.

[viral compass]

Nope, nothing reported at all this side. We have a guy building a big community server as well so I'd assume he would of told me if he had issues 😅

[swift pendant]

We have started seeing some save data fail to migrate since the update today
Resulting in the IslandID changing to a new world no matter what

[junior sparrow]

And it worked on a fresh server for me too

[narrow plaza]

Are you guys using the old world location or the new one?

[junior sparrow]

There's a new location?

monka

[junior sparrow]

maaaaaaaaaaaaan

time to make a PR

[swift pendant]

It automatically migrated it into the new folder

[narrow plaza]

Yeah, it did on our end too. Just wanted to double check you didn't have anything panel side forcing the old location.

[swift pendant]

Seems the client isn't even able to load into that save in singleplayer
But it's only happened on about 3-4 servers so far

[viral compass]

They really should remove the Nitrado server support and GPortal support channels. They both gave up helping in here a while ago now 🙁

[errant flower]

My server only crashes when going into the boss battle with israel hands

[junior sparrow]

No reports outside of the single weird connection issue I'm digging into right now.

[prime badger]

for any still having issues, if you select the Force Relay Connection when connecting by invite code, it will connect you, we had 6 join this way seemlessly back to back all our data was still here

[rich dawn]

Damn no issues for about 2 hours then it dumped us again

[opaque spruce]

my server is having "unknown functions" errors and quitting. Any information on this error?

[smoky sentinel]

Peculiar.. Reproduced this using Wine 11.0 🤔

[old folio]

Won't lie, I haven't seen any reports of these issues either, also using wine 11.0

We do recommend our users use direct IP, I don't know if most people use that and it's fine or what may be going on.

Aside from only save uploads - there seems to be an issue in some way when uploading V2 saves per #🚧│tech-support, however I did upload an old V1 save copy for someone and let the server migrate it, that worked fine.

[smoky sentinel]

Yeah it seems to be a weird image side issue, on a goose hunt rn lol

[smoky sentinel]

Okay I think we fixed it lol

Atleast the one method we could reproduce anyway

[old folio]

Can’t help but say I’m curious as to what the issue may’ve been.

[smoky sentinel]

From what I can gauge, this somehow blew up the buffer we have on stdout and thats about the nitty gritty

Like specifically when terraforming and doing other very specific actions; even though literal log output is.. basically nothing lol

Never ran into it before, but thankfully after figuring out terraforming would trigger it; ezpz fix 🤣

[old folio]

interesting.

[smoky sentinel]

Yeah no idea But atleast its fixed! (I hope, anyway)

If not im around all night so 😭

[old folio]

🤞

[viral compass]

Glad you got it sorted. I did a check in with my guys before I went bed and none reported any issues (except one who hadn’t restarted for days and was getting some memory leak lag spikes, but he has a restart schedule now)

[viral egret]

Hey, are any of you seeing this on Windrose containers (Linux + Wine, UE5/R5)?

Pattern we keep observing:

Multiple servers on the same host stop within a 10-60 s window
No Fatal/Assertion/appError in R5.log, no Sentry/Crashpad upload, no stack trace
Host stays up (no reboot, no OOM events in journald, no node_exporter gap)
Container exit looks clean: RAM drops sharply at the exact second the Gameserver stop, no spike before
Exit code 143 (SIGTERM) on the docker container, not 137 (SIGKILL/OOM)
Last R5.log lines are mundane (normal BL traffic, ReportUsageHistoryList), then silence.
Reads like the game backend is telling them to shut down. Anyone seen this on their fleet, or is it just us?

[narrow plaza]

We were experiencing issues like this. I believe Kringle was able to get it fixed. #📡│hosting-providers-chat message

@gentle wind Test the server today when you can. 🙂

[gentle wind]

@narrow plaza Will do later

[narrow plaza]

@prime badger If you are still having any issues may want to look into this. #📡│hosting-providers-chat message

[prime badger]

My server leveled off after we did the force relay, no issues to my knowledge?

[prime badger]

Yea game breaks when terraforming it gets brain dead trying to render in adjustments across multiple players

[quasi oyster]

We're still having server issues on our dedicated servers
Using docker on linux with wine to run it
I've tried completely redoing one today with a fresh server but copying in the latest world backup that was working + ServerDescription.json from the previous server

We are seeing an issue where it works fine to start with and then when users connect it sometimes seems to do something that triggers horrendous stdout spam on the server and after that point no player can join (returned to menu on joining)

stdout is spammed with entries like this:

windrose  |         Id: "EB6DBFE0FE824B986AA0A69467C2E32C|I|19|P3|v42|v0|40|GA|7"
windrose  |         Data: "\000\000\000\000 \000\000\000EB6DBFE0FE824B986AA0A69467C2E32C\000\000\000\000\000\000\000\000\004\000\000\000"
windrose  |       }```
(note: similar entries repeat and don't seem to stop)

once the server is in this state no further players can join. Sometimes a restart wioll help, other times it will not.

I am happy to provide logs and any extra info to the devs if you let me know how to submit it 🙂

[smoky sentinel]

This was reported to their team already after a long amount of head-bashing by me yesterday

On your end you should be able to handle the stdout pipe cleanly and itll stop happening, but its likely hitting a buffer limit

#1500810776150278258 message<

[prime badger]

it was just working. who broke it again?

[smoky sentinel]

?

Did somethin go kablooey?

[prime badger]

we were in game and randomly wiped our server. i deleted the new backups and relaunched.

[prime badger]

yea it randomly wiped our entire server

[prime badger]

it randomly did what mikul's did.

idk what they did. but this has completly ruined the experience for us hours of gameplay and building gone in the blink of an eye....

i'm assuming whatever the cloud auto save is overrode my data and we now have a blank map which is infuriating because cloud backup is not necessary when it is being ran from a dedicated server host with 8TB of storage.

[smoky sentinel]

What host are you on? 🤔

Or selfhosted

[prime badger]

selfhosted

[smoky sentinel]

mm yeah seeing some weird handling w/ saves based on the logs on my test server. For example the check it makes against SaveProfiles shows Worlds: 0 even if it loads the world that exists

[prime badger]

selfhosted on a 10GB Fiberline with 64GB of ram on a dedicated SSD

[smoky sentinel]

Did it crash -> restart -> save is wiped? I havent seen reports of that specifically

[prime badger]

yea and even if i delete the new ones it overrides after finally connecting

[smoky sentinel]

Internally we noticed if you copy the save data to the RocksDB path and the v2 path; it tends to load it fine - may be worth tryin that? 👀

[prime badger]

we randomly couldn't interact or move. then server crashed. i will have to review the logs

[smoky sentinel]

The crash itself is more likely from the stdout buffer if I had to guess, any form of stdout handling or it not having a real terminal can cause it iirc

but the save "wipe" feels unrelated

atleast atm

[prime badger]

it keeps populating a new seed

i have been hosting this server for nearly a month no issues.

luckily i made a backup when they first started having issues a few days back. i'd rather loose a little progress than all of it

firing the hardbackup, wish me luck i suppose lol

lost a lot of progress but i got it back

[prime badger]

this didn't fix it. what did fix it was backing up my ServerDescription.json and my saved folder yesterday morning

i unpacked that into my dedicated server and relauched it and everything is back from the point of <t:1777909320:s> so could have been worse

i heavily advise anyone else to do the same. make a seperate folder eslewhere of the saved and serverdescription in your R5 folder

in the event your server "wipes" drop the good files in to replace and it will bring your world back

[molten sierra]

Hey hey, I'm running a hosted dedicated server by gportal and my Friends started to have lags. I could find that I have a warning in my logs that appears pretty often. Does someone know what's causing this warning?

[2026.05.05-18.44.47:334][203]LogNetPackageMap: Warning: FNetGUIDCache::SupportsObject: BoxComponent /Game/Maps/GYM/Genlandia/GenlandiaMulty.GenlandiaMulty:PersistentLevel.R5POIBlockingVolume_2142268959.BoxComponent_1 NOT Supported.

"...Volume_xxxxx" changes

[quasi oyster]

Hmm thanks, I'll try and do something on this! Have you just piped stdout to null or something?
It feels like it's maybe a number of stdout writes in a given time that causes it as it seems to occur more with players joining close together or high player counts so this feels like it's feasible as the cause of our issues to me 😅

[viral compass]

So that looks like a bug in the map, basically an invisible wall or zone is glitching out on the network side.
It could theoretically cause lag if the system is constantly trying to process and write that peice into the logs, more likely reason is lack of available resources.

Platforms like G-portal operate on a "shared hosting" model. This means your game server shares its CPU, RAM, and disk speed with many other servers on the same physical machine.

You can try submit a ticket to G-portal and ask if they can migrate your server to a less congested hardware node. Sometimes this frees up enough processing power to handle the unoptimized areas of the game.

[molten sierra]

Thanks a lot, I‘ll see if those lags remain and if so I‘ll submit a ticket

[molten sierra]

I'm recently recieving this error while booting the server
[2026.05.06-14.21.05:804][881]LogWindows: Error: appError called: Fatal error: [File:D:\Source\Build\work\gameRepoCheckoutDir\Source\R5\Game\GameStates\Components\R5GameProblems.cpp] [Line: 211]
[000881] UR5GameProblems::UpdateGameStatus Server will be stopped due to data inconsistency [D:\Source\Build\work\gameRepoCheckoutDir\Source\R5\Game\GameStates\Components\R5GameProblems.cpp:211]

[remote meadow]

verify game files

[molten sierra]

Not working

seems to be an issue with Windrose+ and increased XP Modifier

[prisma sorrel]

Seen someone have that the other day. They had to rollback their player data.

[shy heron]

Hello, did you find a solution to get it working?

[tight current]

Thank you for this picture. I spent too long trying to load my back up.

[finite needle]

is there any mod or anyway to stop people from griefing buildings on servers

[prime osprey]

Hey guys, running a remote connect set up. When our server booted up, its starting a new world instead of loading our saved world. Checked the world descript json file, its got a diferent world to ours.

made a back up of saves elsewhere, deleted new folders, changed json to launch old world name, relaunched, it just creates a new world and over writes the json with new world name.

Any iodeas??

had same issue on Aska 3 months ago and bisect gaming staff couldnt fix it either.

[raven dagger]

Probably there's a secondary wine file opening up on restart every time you need to add in nano if im not wrong that it kills that process always

Because that fixed it for me

[prime osprey]

A wine file?

I’m not super techy. I’m winging it. Fixing physical internet lines is more my mainstay 🤣

[molten sierra]

is someone of you having troubles using the windrose+ dashboard?

[raven dagger]

Neither i'm you should try to use chatgpt to troubleshoot got mine fixed 😄

[rustic depot]

Is the high write rate even when no one on the server still a thing? Shut my down whilst no one playing before it kills my SSD

[smoky sentinel]

Heya!

To confirm, is this Windrose server also @ Bisect?

For Windrose, the most typical cause of this is the server trying to load from the Backups dir > RocksDB_v2 - We made a post about this here after our Integration team spent all day debugging it 🤣 -> https://discord.com/channels/1260209458924224542/1501430959701098507

Issue post is for the Config specifically, but the same workaround flow works for loading save data; thankfully!

The TLDR workaround is
-> Rename RocksDB_v2 to RocksDB
-> Move the RocksDB_v2_Backups dir somewhere else on the machine
--> If on a GSP, ensure their Startup / Config as needed on the panel is tied to the Island ID (If you're on us, this shouldn't be needed but also fine to do via the Startup Tab
-> Boot

The unfortunate downside to this messy workaround is map discovery and whatnot may get reset - though it should get the save loaded + ensure it stays loaded!

To touch on the ASKA one, that one unfortunately is kinda on the ASKA devs; and kinda a handle issue on our end. To TLDR this as its another games bug
-> ASKA on read looks for a savegame_id, if this resolution fails in any way (Silent errors, or anything else) the server by force loads a new world; instead of logging the error or exiting
-> The way our end can affect this, is we try to manually upkeep the config for you by reading the Save Directory for the most recent save file. ASKA itself can however still ignore the fact we're doing this and do what it wants (And its been reported, though that was quite some time ago so likely still on their Roadmap if its coming up still)

[old folio]

Aska issue has a fix (atleast that one does)

Can share if curious

[smoky sentinel]

Oh, please do 👀

[old folio]

Line endings.

Check the default config file you’ll see. If they change the file will generate a new world at startup after a crash.

Can confirm this is a fix as that game still crashes daily - better have detection for that too, and even when we have users that don’t specify a save id in our panel anymore, they’ll still have a single world.

[alpine lion]

when i copy my single player world file to my nitrado windrose server and then select the world file in my general tab all my buildings and progress are gone

how do i fix this

[silver cove]

i have the same problem here. Tryied everything i could find online but cant get it work by now

[silver cove]

i found a working solution 😄 I made a complete new install of the nitrado server and then followed the TLDR workaround above in the Chat (renamed the RocksDB_v2 to RocksDB and deleted the Backup_v2)

[glad hearth]

Where to find nitrado support here?

[alpine lion]

i see ill check that out

thank you

[alpine lion]

everytime i delete the backup_v2 it comes back😅

[alpine lion]

and the server wont stop restarting for some reason

[viral compass]

You will have to go to Nitrado and open a ticket, they kinda stopped giving support in here ages ago

[warm night]

Does anyone know which file contains the game settings, such as enemy spawn rate, ore spawn rate, etc.?

[viral compass]

WorldDescription.json There is no control for spawn rates, enemy or ore at the moment, spawn rates and overall hostility are tied to the broader difficulty settings and multipliers in that file.

[lethal sandal]

how to do manual saving on the 24/7 server hosting

bcs if smth happened and made the server shut down, the progress would be gone...

[ebon anchor]

There is no way to manually save right now.

This game has no remote commands or in-game commands. It doesn't even have a chat. It's mainly people relying on third party mods right now for doing things remotely. Only thing I can say is to make backups each time your server restarts.

[lethal sandal]

thank you!

[ebon anchor]

New server setting CanLaunchMultipleServerInstances.

[floral sequoia]

[bronze phoenix]

hey everyone! I’ve been working on a Windrose dedicated server platform and I’m looking for 2 people to try it out.

It’s aus-based (g’day mate), 4 slot servers, and the whole web control panel is built around Windrose specifically (not just a generic host with Windrose slapped on it). big focus on world backups and restores (without ftp, trying to make it more seamless and pain free).

Still early days so I’d love some real feedback from people actually playing. I’m after 2-3 people who want a 4-slot Windrose server for free. DM me if interested!