import centos VM from vmware

[taich]

Hi,

I want to import a centos (oracle linux) 7.9 virtual machine from vmware into proxmox.
I did the export with

"ovftool vi://server/vm destdir"

I also can import the vm with

"qm importovf vmnr vmname.ovf storage"

But I cant get the vm running.
As soon as I start the imported vm I get this error.

dracut-initqueue[313]: Warning: Could not boot
dracut-initqueue[313]: Warning: /dev/mapper/ol_oracle-root does not exist
dracut-initqueue[313]: Warning: /dev/ol_oracle/root does not exist
Starting Dracut Emergency Shell...

Does anyone know how to get this vm run?

 

[Dominic]

Could you post your VM configuration?

qm config <vmid>

Depending on operating system and previous settings, trying some combinations of BIOS/UEFI and IDE/SATA/SCSI for the VM disk can be necessary.

 

[taich]

Could you post your VM configuration?

qm config <vmid>

Depending on operating system and previous settings, trying some combinations of BIOS/UEFI and IDE/SATA/SCSI for the VM disk can be necessary.

qm config 280
boot: order=scsi0
cores: 2
memory: 4096
name: oracle.datamagic.at
net0: virtio=B6:8E:92:76:F9:E4,bridge=vmbr0,firewall=1,tag=1224
ostype: l26
scsi0: DRBD2:280/vm-280-disk-0.raw
scsihw: virtio-scsi-pci
smbios1: uuid=de35d7ec-8fe3-422c-aa94-1f34edc1141f
vmgenid: 91c63c57-e1bb-4dd8-82bc-d07eb0f1dd85

And before export I checked the system for virtio drivers

grep -i virtio /boot/config-$(uname -r)
CONFIG_BLK_MQ_VIRTIO=y
CONFIG_VIRTIO_VSOCKETS=m
CONFIG_VIRTIO_VSOCKETS_COMMON=m
CONFIG_NET_9P_VIRTIO=m
CONFIG_VIRTIO_BLK=m
CONFIG_VIRTIO_BLK_SCSI=y
CONFIG_SCSI_VIRTIO=m
CONFIG_VIRTIO_NET=m
CONFIG_VIRTIO_CONSOLE=m
CONFIG_HW_RANDOM_VIRTIO=m
CONFIG_DRM_VIRTIO_GPU=m
CONFIG_VIRTIO=m
CONFIG_VIRTIO_MENU=y
CONFIG_VIRTIO_PCI=m
CONFIG_VIRTIO_PCI_LEGACY=y
CONFIG_VIRTIO_BALLOON=m
CONFIG_VIRTIO_INPUT=m
# CONFIG_VIRTIO_MMIO is not set
# CONFIG_RPMSG_VIRTIO is not set
CONFIG_CRYPTO_DEV_VIRTIO=m

 

[Dominic]

You can try IDE or SATA for the disk. Are other VMs working with the DRBD storage?

 

[taich]

You can try IDE or SATA for the disk. Are other VMs working with the DRBD storage?

I tried SATA and IDE but nothing changed.
The virtual machine provision setting in VMware is

SCSI controller = LSI Logic Parallel
SCSI Controler = SCSI (0:0)

This is the original disk layout in vmware

# lsblk
NAME               MAJ:MIN RM  SIZE RO TYPE MOUNTPOINT
sr0                 11:0    1 1024M  0 rom
sda                  8:0    0   60G  0 disk
├─sda2               8:2    0   59G  0 part
│ └─ol_oracle-root 252:0    0   59G  0 lvm  /
└─sda1               8:1    0    1G  0 part /boot

 

[taich]

You can try IDE or SATA for the disk. Are other VMs working with the DRBD storage?

Yes, DRBD storage is ok, there are several other vm's running on it.

 

[Dominic]

Yes, DRBD storage is ok, there are several other vm's running on it.

Have you imported those from VMware as well or is that oracle VM the first one?

 

[taich]

Have you imported those from VMware as well or is that oracle VM the first one?

I have imported some of the others from the same VMware server. Some of the vm's on proxmox I just created newly. All the other vm's I sucesssfully imported were debian based. I think this is somehow related with the oracle partition layout.

# lsblk
NAME               MAJ:MIN RM  SIZE RO TYPE MOUNTPOINT
sr0                 11:0    1 1024M  0 rom
sda                  8:0    0   60G  0 disk
├─sda2               8:2    0   59G  0 part
│ └─ol_oracle-root 252:0    0   59G  0 lvm  /
└─sda1               8:1    0    1G  0 part /boot

 

[taich]

I also tried to change BIOS but no sucess. centOS does not boot.

Starting Dracut Emergency Shell...
Warning: /dev/mapper/ol_oracle_root does not exits
Warning: /dev/ol_oracle/root does not exist

 

[PTS]

Hi all,

Being a newbie on Proxmox (but with strong background on VMware) I found a couple things that need hacking when moving [linux] servers from VMware vSphere to Proxmox VE with ovftool.

1) Switch your vm disks from behind SCSI controller to SATA on VMware already. Reboot to see it works before migration. Install qemu-guest-agent before final shut-down.
2) If your server has mutiple dots in name (say, server1.domain.network) the qm omportovf always fails. To get it working, you'll have to rename all related files (e.g. to server1-domain-network.ovf, .mf, -disk1.vmdk etc.) and do "sed -i 's/server1.domain.network/server1-domain-network/g'" on your .mf and .ovf -files. Then qm importovf
3) your network adapter is gone. Period. You have to create new one. Most sure-fire inside linux is to use NetworkManager (and nmcli) instead of wicked, but that's also up to your taste and liking of linux administration.
4) when your server starts okay on Proxmox, you can power off, change network adapter to virtIO and chipset from i440FX to q35 and power on.
5) If you try EFI and return to BIOS, the EFI disk is left there. This blocks live migration. Finalize EFI-configuration or remove unused EFI-disk to get live-migration working again.

That's first 5 from me. Anything you'd like to share?

 

[ness1602]

Can you get to recovery mode and rebuild all with : dracut -f --regenerate-all ,and try to boot?

 

[taich]

I did that but it did not help.
In the meantime I gave up and installed a new vm.

 

[yohighnest]

Its an old thread, but i had the same issue. i managed to get it running to but into the system via grml/chroot and recreate the initrd with all drivers. the kernel version is the one from /lib/modules

dracut --force --no-hostonly '' your_kernel_version"

 

[mrtt78]

i've had the same issue,
solved following this istructions: https://blog.bianxi.com/2022/09/18/convert-oracle-linux-7-9-to-proxmox/

i was just having the VM with this kind of settings,

• BIOS: SeaBIOS (Should be able to see Grub menu)
• Machine: Default (i440fx)
• SCSI Controller: VirtIO SCSI (It might not be used as sata0 to be considered for disk)
• Hard Disk (sata0): disk_image_file
• Network Device (net0): vmxnet3=<mac_address> (This is default for VMware, can use other type too)

and disk imported in qcow2 format.

I've been able to boot in rescue mode, discorver my last uek kernel version with "rpm -q -a | grep kernel | sort",
find scripts in kernel package "rpm -q kernel-uek-<version>.el7uek.x86_64 --scripts"
and launch the two commands in the previous scripts like
/usr/sbin/new-kernel-pkg --package kernel --mkinitrd --dracut --depmod --update <version>.el7uek.x86_64
/usr/sbin/new-kernel-pkg --package kernel --rpmposttrans <version>.el7uek.x86_64
to rebuild grub, then reboot.