Jwoff11

Gotcha! A little sleuthing online shows that this file is notorious for not being generated. Maybe I can look into a way to manually force it or have a generic copy in the repository that downloads there? Food for thought but understand you aren't looking to address that. For me being new to Unraid, why do you have to delete the container variable for it to apply correctly? Setting the variable to something different doesn't automatically change the port number on the start-up line? If it doesn't then why have the variable and just make sure it's in the startup line? Just trying to understand and learn more! I was thinking more of a variable that allows you just to enter the list of ModIDs and then it's job would be to add it at the end of the extra startup params when starting the server. That way mods have their own line that's easy to manage. Ultimately, not a big thing either way and I understand why you wouldn't want to support that. While it was difficult for me to set up, this is easy to do now that I know what is needed. If the container has a default volume mapping variable that allows for a user to enter a folder share location on a pool or array, the only other items needed are a -cluserid variable and a -ClusterDirOverride variable for the user to populate. See an example I have done for the container below:For any server that is going to be clustered together, this is fine and does not need to be different because they will be sharing the same file to keep track of items, dinos, and characters that are uploaded and traveling between servers. So if there was a boolean added for if the server is going to be part of a cluster or not, you could just have three variables be looked at if true. Those three being the -cluserid, the -ClusterDirOverride, and the volume mapping needed to send the directory file to a share the servers see. Lots of words, but if I need to explain better let me know. Ultimately though, if you don't want to, can't make you. Just saying it'd be easy. Yeah, this one was me using the wrong startup command....whoops! Working now.... Initially, this intimidates me a bit, but maybe if I get some gumption I'll give it a try. Might be PMing you for help though! I will be working on this in the coming days. I had to do so much research to get this one going and I would love to save people some time. Appreciate the help and the insights!

@ich777 just to let you know, I successfully got a modded, clustered ASA up and running! I have a few requests/observations to hopefully make it easier for other users to get it spun up. And with less questions about it posted here 😁 The Game.ini file is not automatically generated inside the same folder as GameUserSettings.ini(Default Location is /mnt/user/appdata/YourArkContainter/ShooterGame/Saved/Config/WindowsServer). This file is pretty important if you want to fine-tune your server's settings. If the file is manually put there, the server will see it(After some tweaking which I will cover below), but it would be nice if it was automatically generated. I have read that sometimes this is just the Ark Server files being clunky but just thought I would bring it up. Port 7777 is the default port for the servers, but it seems like the UDP variable has to be deleted and readded with the different port number for it to actually update to the new port number. Could this be something I am doing wrong? Is there a way to add a separate line for the Mods list? With it being in Extra Game Params, it can be difficult to edit with a long modlist on one line with all the other params. All the server is looking for is -mods= and then the modIDs. Maybe it could be made easier by allowing a copy-paste list in a variable or have it look for a txt file that has all the modIDs. Spit-balling here, but I think the main thing is just making it easier. It works now, but having a dedicated variable in the container might help people. Having some built-in logic for Clustering would be awesome. For Clusters to work you have to have the -clusterid= and the -ClusterDirOverride= variables in the extragame parameters. But what you also have to do is set up a volume map from the container to a folder location on your Unraid server(assuming you're running multiple containers on one machine). Then you have to set the volume-map inside the container as the variable for the ClusterDirOverride. (For Example: /ark/clusers --> /mnt/user/appdata/Clusterdirectories, but the variable on the start line would be -ClusterDirOverride="/ark/clusters"). Since it is just a container volume map, could you add to the template a boolean option to cluster the server or not? Then, if true, check the mapping variable for the user's assigned folder location that the mapping will point to. If the folder is not there the container could generate that folder, assuming PID and GID are left the same in the container variables. Then you could add a variable for the -clusterid that is based on that same boolean value being true. That way those commands are always on the startup line unless the boolean is set to false. Also I don't think if these are clustered together, that it would need a separate steamcmd, but I could be mistaken here. But if I am right, it might save on download time if there was a separate steamcmd directory it could point at and use for all the servers instead of having to download it each time for each server.(I am still learning here so I could have a misunderstanding of how this works) I mentioned that Game.ini and GUS.ini needed some tweaking to get it recognized by the server. What I have noticed(And still don't have a great workaround for) is the container seems to use the previous values of the game.ini and the GUS.ini unless portions of the GUS.ini file are missing at startup. Then it will read the whole file and apply settings. The only way I have figured this out was in trial and error of getting the settings updated. And to further compound the problem, sometimes even doing that process fails to work. My request is can you have the container check and update the server based on the GUS.ini and the Game.ini file every time the server starts? It seems like at the moment it is not working reliably. That's all I have for now. If you could make those updates, I could write out a workflow for how to get modded clusters up and going. And I might have it wrong on some areas, so let me know if I do. Really enjoying doing this and learning more about how this OS/Container works!

Honestly, I am going step by step to see what I can do. Learning a lot by just getting my hands dirty. Going for a cluster now and the only thing I can think of is multiple containers on different ports but have them have the same cluster ID.....once I can hop between maps, then I'll try Mods....

I appreciate you replying. I have read many of your replies here and know you deal with many questions. Based on your knowledge, do you know if people have successfully run mods in this container? I looked on this thread and around online and it seems mixed.

@Cyd would it be possible to use this in conjunction with @ich777's ASA Container? I know this is meant for ASE, but hopefully, it would work the same.

It does fill completely and I have to restart or move the data location for the image to empty.

I can google this, but if you had a guide or reference material on how to do this it would be greatly appreciated.

wouldn't that still be an issue thou? I mean I guess I could setup a cron job to restart docker to empty the image. Not sure if it's possible to do that or not.....

Ok so the docker image is still filling up after a fresh install and correctly mapping the locations. I was able to do a Docker Run like you asked @trurl and I have attached it here. Any chance it's still something I have configured wrong? Any time I update the docker container, it dumps the image. If that helps you at all.

Was able to get it back up and running. If anyone accidently deletes their MineOS UnRAID container, just copy paste over the files from your old server to a new instance and that should work. Feel free to ask me about this if this doesn't make sense. As far as my docker image goes, I will continue to monitor and report back.

So I got home and the server won't start. It's odd. If I pull a backup from yesterday at 2 am, it runs. But not the files from right before I accidently deleted the container. I am currently making a new MineOS server to run over night to confirm the docker image does not begin filling up

Well I tried updated a Port Range, which caused the docker run command to fail and delete the container. Currently at work so I am not sure of the damage, but when I was remotely installing the container back, I realized that the user folder that was inside the mnt folder wasn't on the cache drive I had selected. Once I went back up and was able to select my cache drive, I believe that fixed it, but I am not sure at what cost. I have backups of my world so it should be ok. I'll report back later tonight.

Also wanted to add this. It's a script from a SpaceInvaderOne video that checks for volumes on the docker image. Right now, the only thing I see is the container itself is filling up. But I am not sure why?

Here is this. Dunno if it will help or not.

It's still MineOS taking up the space but I have no idea why it keeps filling up if I have the data stored on a dedicated cache drive...