Hello! Sars here. I've been trying to get UMH working on our systems by following the installation instructions documented [here](https://umh.docs.umh.app/docs/getstarted/installation/). According to the documentation, UMH can run on three different operating systems. However, none of the documented methods have worked for us. Our Setup: - AWS EC2 (xlarge - 80GB) - Local Server (4 cores, 16GB RAM, 1TB storage) We launched three separate EC2 instances with the same specifications but different operating systems: Ubuntu 22.04.04, RHEL 9.4, and Flatcar Stable from the Community Store. Executing the script on each resulted in the same error loop. I've attached the logs for reference. After encountering issues with these methods, we decided to try UMH's customized Flatcar OS. Following [this](https://umh.docs.umh.app/docs/production-guide/installation/flatcar-installation/) guide, we used a local on-premise server and attempted to install UMH's customized Flatcar OS. Initially, we faced a problem as BalenaEtcher couldn't recognize the .usb extension. However, using the flash from URL option worked. We got through all the steps until the final screen where you click Confirm [Default]. At this point, the installation failed, saying the Network is Unreachable. We've tried all options except Proxmox, but none have worked. I was considering creating an issue on the Github repository but thought I would reach out here first. If necessary, I can still create a Github issue.

Hi, let me try to help you. First, we will remove flatcar from the community edition as it is currently not really integrated in the self-service product and more of a enterprise only feature. About Ubuntu and RHEL: every version of the UMH is tested automatically multiple times against Ubuntu and RHEL / Rocky Linux, before we release it. So if you encounter any issues, it must be because some deviations from the "base", which could be default AWS settings for permissions or similar. My first question: I am seeing that the logs end after the first k3s installation retry. Did you abort the script or did you trim the logs? We have noticed that k3s requires multiple installation attempts to deterministically install it in a fixed time, as there is a race condition within it that requires a system restart. So as a solution, we remove it and retry again, and so far it has always worked within the limit of 5 attempts.

Hi, Jermuk. Thanks for the reply. > it must be because some deviations from the "base", which could be default AWS settings for permissions or similar. I see. Although, we did try installing UMH on our local server running supported OSs, it again resulted in the same error loop. > Did you abort the script or did you trim the logs? Ah yes. Apologies for the oversight - the logs were indeed trimmed. We left the installer running for a couple of minutes, but it remained stuck in the same loop of retrying & reinstalls of k3s. I have been doing some testing on local Ubuntu 22.04 server, and I have observed that when you've k3s preinstalled, the installation succeeds.

did you let the loop finish? it willautomatically close after the 5th retry. and sometimes it really needs 4 or 5 attempts...

The loop never finishes... it retries 5 times, uninstalls k3s and repeats. Is there also an "outer" limit as to how many times it can retry retries?

let us troubleshoot internally. did you get at least one UMH instance running so you can get started with playing around?

Yes. I have been able to get it working by having k3s pre-installed