if its just a few friends maybe use tailscale? put it on the server, then have your friends run it so their on the same network. then use tailscale internal IP’s and dont worry about DMZ or port forwarding?
Im running mine on unraid but as they mentioned, Just make a VM in unraid and install AMP on that. Linux/ubuntu is recommended. Works well! (I went with windows server because space engineers is my main gig but there are caveats to using windows)
After running this nearly a week without problem . Today I rebooted the host and now this is what showing
well after shutting down the server for last 12 hours. Powered it up today and updated amp and the instances. it started just fine.
now when I go to deep dessert I see 2 instances popup.
but when going into it it shows this:
[00:32:11 11 DBG RMQA] TravelQueue DeepDesert_1 received message: {
"MapName": "DeepDesert_1",
"IssuedUnixTimestamp": 1780187531,
"Players": [
{
"Id": "55F71A1EA3070779",
"TargetDimension": 0
}
],
"OriginId": "Overmap2",
"RequestID": "658F12068A0C4F92AA5C12F42FF579DC",
"PasswordOrToken": ""
}
[00:32:12 16 INF Main] Received travel request for 1 player(s) to DeepDesert_1 (instancingMode=Dimension)
[00:32:12 16 INF Main] Notified player(s) "55F71A1EA3070779" of travel response Overmap2: {"Code":3,"OriginId":"Overmap2","RequestID":"658F12068A0C4F92AA5C12F42FF579DC","MapName":"DeepDesert_1","DestinationPartitionId":null,"QueueToken":null,"QueueState":null,"ServerState":null,"BroadcastExchange":null,"ServerFull":false,"ServerLoginToken":"","RetryTime":null}
[2026.05.31-00.32.12:216][121][1472][2]LogTravel: Error: [658F12068A0C4F92AA5C12F42FF579DC]: sb7Q2$ Invalid dimension
[2026.05.31-00.32.12:216][121][1472][2]LogTravelEvent: Error: [Failure] FlowType:"Travel", Stage:"Request", PlayerId:"55F71A1EA3070779", FlowId:"658F12068A0C4F92AA5C12F42FF579DC", Reason:"TravelVolumeOverland", ErrorMessage:"sb7Q2$ Invalid dimension"
So I'm assuming Mike must be doing something.
@blueSomnambulist I see you got the final harvest working. I am using the standard self-host server funcom gives you on steam nothing changed. I’m super new to all this, how can I get this working? I cant progress the story or use dual blades till i can get this working. I truly hope you can help me as this is the only mention anywhere on the net about this issue not working.
I’ve only used the AMP version, which uses mock-k8s, which i assume is an emulated Kubernetes cluster. I don’t know how your steam version works, either it is a full Kubernetes cluster or uses the mock-k8s as well. Either way there should be some equivalent of setting the replicas value of the dlc maps to 1 (as described in my third post) which would startup the UE5 servers for those maps. This will cost additional resources and only holds until the next server restart. If your version is running an actual Kubernetes cluster it’s possible that setting EnableAutomaticInstanceScaling=true will work and the server will automatically start the instances once you try to go there. That didn’t work in the AMP version for me.
I’ve managed to get the game server to be stable trhough normal gameplay, visited the main cities and Deep Desert, etc. but I’m having a weird issue ONLY with Dune.
Every single morning when I check AMP my Dune instance is stopped. ONLY my Dune instance. Even without an update/restart task scheduled. Even without a backup scheduled. Even with those events completely DELETED and AMP restarted after deletion, it stops nightly. Nothing in the event log states it’s being stopped or restarted. The console log is spam city, but it doesn’t look like a crash OR a purposeful shut down. I don’t have enough Adderall to read each line of hundreds though.
Any ideas what’s going on or what I could do to isolate the issue a bit to try and solve? It’s not really a BIG deal, data isn’t lost and I just have to hit “start” in the morning, but still.
Ah dang. I tried the setting the scaling to true but that didn’t work. I will reread you third post and hopfully i can get it working. I was excited as I saw the server needed a patch today, sadly the fix was not part of it.
I dont think 32GB is enough anymore with the warm DD map.