Open berendt opened 7 months ago
Tested it with an AMD Ryzen System using Windows 11 - Virtual Box 7.0.20 - I can't get Ubuntu 22.04 to boot when I enable the UEFI Bios at the VM.
9 out of 10 Times it won't boot into the Kernel after Grub Menu, and if it's booting the ISO the subiquity Ubuntu installer will run and partition the discs and coping files, but at a random point it crashes and shut down the VM.
I also tried to download the live cd from ubuntu, but same happed, no boot into kernel after enabling UEFI Bios Settings, legacy bios is working, but then the subiquity Ubuntu installer can't partition the discs like under Linux VirtualBox.
After Cilium Network fix for the install script, tried some more runs with different Virtual Box settings to get a working installtion.
As it seems key to success is enablinge EFI and for the SATA Controller, Host Caching must be enabled, with this Setup the ISO Setup, Boot Straping and Deploy is running with no timeout issues and as fast as possible.
Deploy takes really long but thats the problem of only having 28GB RAM for the Virtual Maschine, but it works, not recommended.
I will update the Issues with a working configuration for VirtualBox.
Created a pr for documentation update about VirtualBox Setup https://github.com/osism/osism.github.io/pull/665
I was able to start the automated installation with VirtualBox 6.1 under Linux Laptop with 32GB RAM and SATA 500GB Disc in the VM on a NVME Drive. (Had no more space left)
Important - EFI BIOS must be enabled for the VM - System - Mainboard EFI activate
If EFI Bios is not enabled and it runs in default BIOS Mode the subiquity Ubuntu installer can't setup bootmanager partiton and will fail with and error and stops the setup right after some minutes. For legacy Bios setup the autoinstall.yaml needs changes at the Stroage: Section as it seems. I think this is the major reason why most installations will fail right at the start.
With this Setup, the Disc get partitioned and after that bootstraping the VM will be shut down - all fine until now.
After the next start, the VM boots up and the bootstrapping process starts.
Then I had the problem on different virtualisation tools that it will stop with an error at the setup of Netbox, as it seems it runs into a timeout as it takes to long on the Virtual Machines to start containers, but Netbox was started.
bootstrap | FAILED - RETRYING: [manager.systems.in-a-box.cloud]: Wair for an healthy netbox service for 60 retries and it stops with a FATAL - FAILED that it unable to get a healthy status for netbox-netbox-worker-1.
After that I started the bootstraping script again, and it worked fine and completed.Now i started the deploy script with the sandbox option and installation works until it stopps with error later on "Install Cilium" with
"task includes and option with an undefined variable. The error was: cluster_idr is undefined.
Restarting the deploy script ends after a longer run with the complete crash of the Virtualbox VM.
With a more powerful Hardware it should be running in VirtualBox, as it seems most of the failures we came up with are timing problems. I will test it on Windows in Virtual Box with a more powerfull PC Hardware.