Boot gets stuck at mulitpath siblings

User Support General
,
1

I am unable to boot into the Qubes iso at all. I have tried two different installation media on my pc and they both have the same issue. I tried the iso on a laptop and it was able to boot so I don’t think the issue is with the drive.

Any help to resolve will be greatly appreciated. I am using an nvidia which I see can be a problem and I don’t have an integrated. Is that the end of it and if I want to use qubes then I must change to AMD?

The boot sequence gets stuck and this is all the output that there is.

Started plymouth-start.serse - Show Plymout boot screen
Started systemd-ask-password-plymouth.path - Forward Password Requests to Plymouth Directory Watch.
Reached target cryptsetup.target - Local Encrypted Volumes.

Reached target paths.target - Path Units.
Started cancel-mltipath-wait-sda.timer - cancel waiting for multipath siblings of sda.
Started cancel-multipath-wait-nvme0n1.-timer - cancel waiting for multipath siblings of nvme0n1.timer
Started cancel-multipath-wait-nvme1n1.timer - cancel waiting for multipath siblings of nvme1n1
2

Try to boot using the Qubes OS with kernel-latest boot entry in the installer GRUB menu.
Try to add nomodeset kernel command line option in the GRUB boot menu entry like this:

3

Thanks! That worked. I’ve managed to install but now resolution is stuck at 1920x1080. I saw something about nvidia drivers being problematic. I haven’t tried anything yet but do you think it will be possible to get higher resolution?

4

You need to use the proper driver for your GPU instead of nomodeset (nouveau or proprietary nvidia driver).
You need to check why the default nouveau driver didn’t work for you. Maybe try to use different kernel version in dom0.

5

Thanks for the tip

I updated to the latest-kernel and now lightdm won’t start. First it got stuck in the same during boot and then when I added nomodeset, lightdm won’t start at all. Linux and nvidia are just no fun

6

Maybe there is a regression in the latest kernel version like this:

You can try some older kernel-latest versions.

7

Can you direct me to documentation on how to install an older kernel-latest?

8

List available package versions:

sudo qubes-dom0-update --action=search --showduplicates kernel-latest

Install specific package version:

sudo qubes-dom0-update kernel-latest-6.7.7

To boot the older kernel version in GRUB menu enter “Advanced options for Qubes (with Xen hypervisor)” → “Xen hypervisor …” → boot entry with specific kernel version.

9

Thanks. I managed to install 6.9.10 but the same problem persists. Lightdm crashes and it seems like a problem that X didn’t start.

10

What’s your GPU model?
Try the kernel-latest-6.7.7 for a test.

11

RTX 3080 12gb

There’s no kernel-latest-6.7.7. Even when I search, there’s a 6.4, 6.5, 6.6 and then it jumps to 6.8.

12

kernel-latest-6.7.7 seems to be from the testing repository. Try the kernel 6.8 then.

13

Same problem as before :frowning:

Not sure what logs I could even provide. I checked the lightdm status and its just an error.

14

You can check the logs in dom0 using journalctl and search for a errors/failures related to it.

15

I checked journalctl and Xorg crashed and dumped core. I’m not sure which part might be relevant, It looks like the nouveau driver is the problem. I’ve attached a screenshot of some of the dump

IMG_6812
IMG_68121920×2560 332 KB

16

Check the first related error messages.

17

This was all I saw about lightdm

IMG_6810
IMG_68101920×2560 309 KB

18

Check the errors before lightdm in journalctl. Maybe there are errors related to nouveau.

19

There’s only another Xorg core dump which is a few pages long. Is it worth sharing?

20

Maybe if you search for the significant strings from the coredump you’ll be able to find the same problem described somewhere else.

Try to use kernel-6.6.36, maybe it’ll work for you:

Maybe there is some incompatibility between new kernel versions and old Xorg package from dom0 (Fedora 37).

1 Like