Note - If you do not fill in every section below, your post won’t be answered - you must provide the steps you have followed so far and the actions you’ve already taken. Make sure to remove this notice from your post too.
Screenshots of text are prohibited. Always copy and paste text as text.
Make sure you search before posting! Duplicate posts for the same issue may not be answered.
OS Name/Version: Windows Server 2025
Product Name/Version: 2.6.3 - 20250926.1
Problem Description: ARK: Survival Ascended has an issue with the deployment configuration - any server that is started using AMP will get a default cluster ID of some value (I haven’t been able to figure out what it is or where it’s coming from) causing users to be unable to download survivors, tames, or items to the server from other servers or local saves as the server is looking for a nonexistent cluster directory. This behavior is verified as DIFFERENT from a standard server install via SteamCMD NOT using AMP.
Steps to reproduce on AMP-managed server
- Create a new ARK:SE instance - you may customize the name and password
- Create a new ARK:SE server not using AMP - IE, manually configure the install (I copied most of the /ShooterGame folder for this step just to have the same game settings)
- Enter into a single player world of yours on ARK:SE and upload a survivor, tame, or item
- Log into the AMP-managed server
- Make your way to an obelisk
- Once there attempt to download your survivor/tame/items - you’ll note there is nothing there to retrieve
- Select the tab to upload your character; you’ll notice the prompt is “travel to another server” which is only seen in clusters. The browser here also shows other servers with no cluster ID specified, many are “AMP Powered ARK Server” which is the default naming for an ARK instance through AMP leading me to believe that many inexperienced users such as myself are either unaware of the problem or unable to correct it themselves"
- Log into the NON-AMP server
- Repeat steps 5 -7. You’ll notice this time that on the upload tab you actually see “upload survivor” which is the expected prompt, and that in the inventory/creatures tabs you’ll now see any of the things you uploaded from your single player or non-dedicated world.
Actions taken to resolve so far:
The ONLY solution that I have found so far is to host a server outside of AMP, transfer all of the tames and survivor/items, THEN wait for an autosave and copy the /SavedArks folder to the AMP instance’s datastore. Not exactly the most graceful procedure. The expected behavior works OUTSIDE OF AMP. It does NOT behave as expected when being managed by AMP.
Can still understand your post without the need for caps, bold, italics etc lol
I suspect I know the reason, tho it’s interesting that this is the first time this has been reported
Given limitations in terms of how amp handles command line arguments, in light of Ark having two formats for them, the template adds -clusterid= as an argument but obv if the setting is not completed it is blank. Rather than ignoring it, it seems Ark treats it as the cluster id being set, but to nothing
The only way of getting around this is to add another setting that specifically “turns on” the cluster settings when someone wants to use it - which will impact existing instances that use clusters. Or remove the settings from AMP entirely - which will affect everyone who does or wants to use them
1 Like
If you are up for some testing to confirm this is actually the cause, I can tell you some changes you can make to your instance
I do apologize if the formatting seems aggressive, I’m quite used to being incredibly explicit with reddit posts 
My only intent is clarity, not urgency or negativity.
I can make changes as needed! Let me know what to try and I’ll do my best.
OK
Stop the instance
Using AMP’s file manager, go into the instance datastore and edit GenericModule.kvp
Remove -ClusterDirOverride="{{ClusterDirOverride}}" -clusterID={{clusterID}} from the App.CommandLineArgs line (should be line 40)
Save and close, then see how you go
You brilliant human you! That has entirely resolved it.
I’m not sure what to make of it in terms of next steps, but at least for me and my purposes just knowing to remove those args and add them back in when I want to cluster worlds is workable. Much more than having to jockey files around two servers.
I’ve got no choice but to add the “enable cluster” toggle. This will be fun …
Well at least I can say I found a quirk that might be affecting a few other users.. 
Cheers for the help with this!
Can you confirm that with the current template setup, you don’t see any “upload” option, only “transfer to another server”?
BTW, the template has now been updated.
As a compromise to ensure existing instance clusters are not disabled, the new toggle is on by default. We will keep it as such for a transition period, then revert to it being disabled by default.
So for your use case, after updating your instance, you will need to disable the toggle.
Yes I can confirm this - sorry I had gone to bed for the night.
The previous template only resulted in the “travel to another server” no “upload”. I can test the new template in about an hour when I take my lunch break!
Edit:
I can now confirm that it is behaving as I would expect after disabling the toggle. I created a brand new instance, loaded in and I’m now seeing the correct option with cluster toggled off.
1 Like
You could have just updated your existing instance, but either way thanks for confirming