[SOLVED] /dev/centos/root does not exist after migrating CentOS7 from vmware to proxmox VE

wong

Member
Dec 21, 2020
8
1
23
25

and imported this into proxmox VE. Upon booting I get the following errors:​


1608512779705.png

1608513041991.png
 
Maybe you could reinitialize initramfs with dracut -f --regenerate-all if i'm not mistaken? You can do that with safe mode,or ISO.
Regards,
 
same error. suddenly we yust reboot proxmox and than this error occurse.
what can we do?
centos7 rescue did not find any disk.
we move the disk to qcow2 format and mount it from the proxmox host (/var/liv/vz/....)...partitions are there and data are there.

btw:
many thanks
in rescue
blkid did not show the disks.
 
Last edited:
ok we change the type of SCSI Controler to LS and back to VirtIO, now with rescue boot we can do
vgchange -ay
but it says refusing activation of partial LV centos/root

reboot centos7 normal still hung on dracut with
/dev/centos/root not found


2021-06-11 11_46_14-cluster26 - Proxmox Virtual Environment.png
 
we got it:
we do
vgextend --restoremissing <volume group> <physical volume>
this works.
confusing because lvdisplay shows no missing PV
 
vgextend --restoremissing <volume group> <physical volume>
@TheMrg I know this is a little old of a thread, but I am running into this same problem, and my CentOS7 machine boots and operates fine in recovery mode. I am trying to follow your solution here, but am hitting a wall. Can you provide more details on how you did this?

Here is the drive layout

sda
sda1 part /boot/efi
sda2 part /boot
sda3 part
centos-root lvm /
centos-swap lvm [SWAP]
centos-home lvm /home

Using the above, when you say <v_g> versus <p_v> are you referring to running the command like this:

vgextend --restoremissing sda1 sda

Or am i missing something?
 
This problem is solved by changing the hard disk mode of the VM through the web
In my case I did this:

* Change hard disk mode to IDE (SATA did not work)
* delete CDROM to avoid issues
* detach boot disk
* edit detached disk -> specify "IDE"
(System now boots)

* Login
* Install qemu-guest agent
Code:
yum install qemu-guest-agent

* Rebuild initramfs using this command (from https://endocs.ksyun.com/documents/5425)
Code:
mkinitrd -f --allow-missing \
--with=virtio_blk --preload=virtio_blk \
--with=virtio_pci --preload=virtio_pci \
--with=virtio_console --preload=virtio_console \
 /boot/initramfs-$(uname -r).img $(uname -r)

* Shutdown

* Edit VM Hardware
* Set SCSI controller to Virtio SCSI Single
* detach boot disk
* edit detached disk -> SCSI

The workstation now boots using the Virtio SCSI Single controller.
 
  • Like
Reactions: rudolfo1967
In my case I did this:

* Change hard disk mode to IDE (SATA did not work)
* delete CDROM to avoid issues
* detach boot disk
* edit detached disk -> specify "IDE"
(System now boots)

* Login
* Install qemu-guest agent
Code:
yum install qemu-guest-agent

* Rebuild initramfs using this command (from https://endocs.ksyun.com/documents/5425)
Code:
mkinitrd -f --allow-missing \
--with=virtio_blk --preload=virtio_blk \
--with=virtio_pci --preload=virtio_pci \
--with=virtio_console --preload=virtio_console \
 /boot/initramfs-$(uname -r).img $(uname -r)

* Shutdown

* Edit VM Hardware
* Set SCSI controller to Virtio SCSI Single
* detach boot disk
* edit detached disk -> SCSI

The workstation now boots using the Virtio SCSI Single controller.
Thank you , work on Centos 8 after moving from ESXi
 
Hi Same problem experienced on CentOS 9 trying to convert the VMWare .vmdk file to qcow2 format
If I change the disk type to IDE as suggested previously, I can boot up OK with the qcow2 file (so conversion has worked successfully).

However the next step was to use the mkinitrd command which is a problem now as mkinitrd has been deprecated on CentOS 9 in favour of the dracut command.

Does anybody know what parameters I need to pass to the dracut command to rebuild the initramfs for the virtio modules?
 
Thanks for following up. I've tried so many things , so I'm not sure anymore.
I think I focused on the reply from mmcarn as it allowed me to boot up the qcow2 image which I couldn't before plus it was more recent than the second comment - there's a lot of difference between CentOS 7 & 9.
I'm guessing that I run the dracut regenerate command in the current image and not in Rescue Mode?