Home / server / Questions / 936437
Accepted
Stonecraft
Asked: 2018-10-21 05:09:45 +0800 CST

Apparantly successful virt-install of Clear Linux cannot boot

  • 772

First, I create a clear linux vm with virt-install

virt-install \
--name CLEAR \
--ram 65536 \
--disk path=/guest_images/Linux_main/CLEAR.img,bus=virtio,size=450 \
--vcpus 8 \
--os-type linux \
--os-variant generic \
--network bridge=virbr0 \
--graphics vnc  \
--console pty,target_type=serial \
--cdrom /media/big-tank-8TB/OSISOS/clear-25720-installer.iso

The installer starts up and I choose the only available device "vda", which I assume must be the CLEAR.img file, so I choose to create a parition there and install (also, not sure why no partitions show, this img is on a zvol that I had previously formatted to ext4) enter image description here

The installer apparently completes successfully.

enter image description here

But after rebooting, I am stuck in Seabios "booting from the hard disk"

enter image description here

I tried changing the drive interface from virtio to ide, but it did not help. Not sure what to try next.

I also tried these instructions from the Clear Linux website with which I was able to start up, but ran into problems connecting via vnc, and more relevant to my question, when I tried to virt-install start the provided .img file, I got the same seabios "booting from hard disk" forever that I got after installing via the .iso. I suspect the problem might be that I need to provide a UEFI file like in the Clear Linux instructions, but I am not sure how to do this when using virsh and and existing VM.

kvm-virtualization

1 Answers

  1. Best Answer

    The problem is indeed that you need to provide the UEFI firmware.

    You will need to add the --boot uefi option.

    Reference: https://fedoraproject.org/wiki/Using_UEFI_with_QEMU#virt-install

    You may also need to update the OVMF firmware on your system. Clear Linux provides an up to date one at https://download.clearlinux.org/image/. They also provide the _CODE and _VARS fd's might need to be updated in the nvram section of your

    /etc/libvirt/qemu.conf

        #nvram = [
    #   "/usr/share/OVMF/OVMF_CODE.fd:/usr/share/OVMF/OVMF_VARS.fd",
    #   "/usr/share/OVMF/OVMF_CODE.secboot.fd:/usr/share/OVMF/OVMF_VARS.fd",
    #   "/usr/share/AAVMF/AAVMF_CODE.fd:/usr/share/AAVMF/AAVMF_VARS.fd",
    #   "/usr/share/AAVMF/AAVMF32_CODE.fd:/usr/share/AAVMF/AAVMF32_VARS.fd"
    #]

    This also might be helpful: https://github.com/virt-manager/virt-manager/commit/d2fffa509efe891eef3f2f70e5688c1d6e3a2d87

    You might be fine just passing that flag. Otherwise I'd try the nvram section. Personally I use their start_qemu.sh script. But I wish you luck.

    • 2