You are not logged in.
Hello, everyone.
I have been using Linux mint for several years now, but I am considering to switch to Arch. I decided to test Arch as a guest OS in Virtualbox before installing on physical hardware. But unfortunately, I experienced a minor hickup. I would really appreciate if anyone could help me have a look at this.
I downloaded the ISO-file from https://archlinux.org/download and successfully booted it up. I followed the installation guide at https://wiki.archlinux.org/title/Installation_guide, except for a few minor deviations of my personal choice. I installed GRUB. Then, I shut down the system (in stead of rebooting directly), detached the ISO-file from the settings in Virtualbox, and then tried to boot the installed Arch Linux.
At this point, the booting process got stuck at «Loading initial ramdisk...». Oh well, according to numerous threads in different forums, a whole jungle of possible bugs and errors could cause this.
However, from my personal experiences in the past, Virtualbox have had issues regarding USB 2.0 and USB 3.0 controllers while booting Microsoft Windows. (For example, you could look at this thread. Up until this point (in the process of installing Arch), Virtualbox had USB 2.0 controller activated. So, out of pure gut feeling, I decided to switch to USB 1.1 controller, and try booting again.
AND IT WORKED SUCCESSFULLY!
This was yesterday. I searched the web, and I could not find any thread discussing this error regarding Arch Linux in Virtualbox.
So, today I decided to start this thread, to share my experience. In order to describe my findings accurately, I figured that I needed to reproduce the error, so that I could see exactly what was happening, and possibly produce screenshots where applicable. Thus, I switched back to USB 2.0 controller again, and tried a new boot. However, now the error has suddenly disappeared, and the system boots successfully without freezing, regardless of which type of USB controller that is currently active.
I don’t get this! No software updates have been made to either the host OS nor the guest OS in the meantime. And the host OS was never rebooted. Yesterday, I tried several consecutive reboots with USB 2.0 without success. It appears that switching to USB 1.1, booting + shutdown, and then switching back to USB 2.0 again solved the problem. (For the record, today I also tried booting with USB 3.0 activated, and this also seems to work.)
Moreover, I am no longer quite sure whether this error is caused by Virtualbox or Arch. I am a little reluctant to installing Arch on physical hardware until I have cleared this up, in case such an error should appear and cause a deadlocked state. I would very much appreciate suggestions, if you have any experience.
Host OS:
Kernel: 6.2.0-26-generic x86_64
Distro: Linux Mint 21.2 Victoria (base: Ubuntu 22.04 jammy)
Desktop: Cinnamon 5.8.4
Virtualbox:
Base software version: 6.1.38_Ubuntu r153438
Extension pack version: 6.1.38r153438
Virtualbox configuration for guest OS:
Name: Arch
Type: Linux
Version: Arch Linux (64-bit)
Base memory: 12288 MB (Comment: The host OS has a total of 32GB physical RAM available, so I believe this configuration should be OK.)
Boot order: Optical (first), Hard Disk (second)
Chipset: PIIX3
Extended features: I/O APIC (enabled), EFI (enabled), Hardware Clock in UTC Time (enabled)
Video memory: 128 MB
Monitor count: 1
Graphics Controller: VMSVGA
3D Acceleration: Disabled
Storage devices, IDE-controller: Optical Drive (Device 0)
Storage devices, NVMe-controller: Arch.vmdk 40 GB Fixed size (Port 0), Home.vmdk 60 GB Fixed size (Port 1)
Network, Adapter 1: Attached to NAT
USB Controller: See the initial summary of my experience.
Commands executed, in chronological order:
loadkeys no-latin1Comment: I am using a keyboard with Norwegian layout.
ping archlinux.orgComment: Successfully verified that the guest OS had a working internet connection. Thereafter, I hit Ctrl+C to terminate the pinging.
timedatectlComment: Successfully verified correct date and time settings for hardware clock.
fdisk -lComment: Found /dev/nvme0n1 and /dev/nvme0n2. The .vmdk-files were both created brand new, so no partitions or tables existed up until this point.
fdisk /dev/nvme0n2
g
n --> 1 --> [Enter] --> [Enter]
t --> 1 --> 20
wComment: fdisk was terminated after writing to disk.
fdisk /dev/nvme0n1
g
n --> 1 --> [Enter] --> +300M
n --> 2 --> [Enter] --> +2G
n --> 3 --> [Enter] --> [Enter]
t --> 1 --> 1
t --> 2 --> 19
t --> 3 --> 20
wComment: fdisk was terminated after writing to disk.
mkfs.ext4 /dev/nvme0n2p1
mkfs.ext4 /dev/nvme0n1p3
mkswap /dev/nvme0n1p2
mkfs.fat -F 32 /dev/nvme0n1p1
mount /dev/nvme0n1p3 /mnt
mount –mkdir /dev/nvme0n1p1 /mnt/boot/efiComment: I do realize that this last command deviates slightly from the installation guide. However, I took inspiration from where the EFI-partition is mounted in Linux Mint, so I believe it should work. If you believe otherwise, please explain.
swapon /dev/nvme0n1p2
mount –mkdir /dev/nvme0n2p1 /mnt/home
nano /etc/pacman.d/mirrorlistComment: Added my preferred mirror to the top of the list. Then saved and exited the file.
pacstrap -K /mnt base linux linux-firmware dosfstools exfatprogs e2fsprogs sof-firmware networkmanager nano man-db less man-pages texinfo pacman pacman-mirrorlist pacman-contrib
genfstab -U /mnt >> /mnt/etc/fstab
arch-chroot /mnt
ln -sf /usr/share/zoneinfo/Europe/Oslo /etc/localtime
hwclock –systohc
nano /etc/locale.genComment: Uncommented the lines for “LANG=en_US.UTF-8” and “LANG=nb_NO.UTF-8”. Then saved and exited the file.
locale-gen
nano /etc/locale.confComment: Added “LANG=nb_NO.UTF-8” to the first line in the file. Then saved and exited the file.
nano /etc/vconsole.confComment: Added “KEYMAP=no-latin1” to the first line in the file. Then saved and exited the file.
nano /etc/hostnameComment: Added a hostname of my choice to the first line in the file. Then saved and exited the file.
passwdComment: Set a password of my choice for the root user.
pacman -S grub efibootmgr
grub-install –target=x86_64-efi –efi-directory=/boot/efi –bootloader-id=GRUBComment: This last command deviates slightly from the installation guide. This is due to that the EFI-partition, in my case, is mounted at /boot/efi, and not at /boot which I suppose would the most commonwide practice for Arch.
grub-mkconfig -o /boot/grub/grub.cfg
exit
umount -R /mnt
shutdown -h nowFrom here, after detaching the ISO-file, the boot sequence got stuck during my next boot attempt.
Offline