Restoring 32bit Windows7 VM Saved State Fails => fixed in SVN/next maintenance

[morr_rva]

VM boots cleanly, and creates snapshot without problem. When restoring VM from a saved state snapshot, it will always fail with the following:

Failed to load unit 'lsilogicscsi' (VERR_SSM_LOADED_TOO_LITTLE).

Result Code: NS_ERROR_FAILURE (0x80004005) Component: ConsoleWrap Interface: IConsole {872da645-4a9b-1727-bee2-5585105b9eed}

Guest Additions are current. I have recreated snapshot without any effect on the problem. This worked correctly with the same VM under 5.2.22 and before.

[Socratis]

1. Did you post a similar question recently in the forums?

2. Can you try to switch from SCSI:LsiLogic to SATA:AHCI? Does the "Restore snapshot" error go away?

[morr_rva]

Replying to socratis:

1. Did you post a similar question recently in the forums?

2. Can you try to switch from SCSI:LsiLogic to SATA:AHCI? Does the "Restore snapshot" error go away?

No, I haven't posted a question. I searched for other issues like this without success, though.

If I switch controllers, the VM will not boot at all. I no longer have installation media for such an old OS, and the drivers for that controller are not installed.

[Socratis]

Replying to morr_rva:

If I switch controllers, the VM will not boot at all. I no longer have installation media for such an old OS, and the drivers for that controller are not installed.

Windows 7 an old OS? You haven't tried a DOS VM, right? :)

And you shouldn't need to replace the controller directly and try to boot from it:

• Add a 2nd controller, the SATA:AHCI one. Win7 has built-in drivers, it should find it and install the drivers, no "external" or "installation" media required.
• Now remove the disk (VDI) from the SCSI controller and add it to the SATA one.
• Boot the VM to make sure that everything is in order.
• Remove the SCSI controller altogether.

BACKUP FIRST!!!

[MmnOnZeNet]

I got the same issue with Host windows 10 64 bits and guest Oracle Enterprise Linux 7 64 bits. I attach the logs

[Socratis]

@MmnOnZeNet

00:00:02.510433 [/Devices/lsilogicscsi/0/Config/] (level 4)
00:00:02.510434   Bootable <integer> = 0x0000000000000001 (1)

Yeah, you have what seems to be the same SCSI:LsiLogic issue as well...

[morr_rva]

Replying to socratis:

Replying to morr_rva:

If I switch controllers, the VM will not boot at all. I no longer have installation media for such an old OS, and the drivers for that controller are not installed.

Windows 7 an old OS? You haven't tried a DOS VM, right? :)

And you shouldn't need to replace the controller directly and try to boot from it:

• Add a 2nd controller, the SATA:AHCI one. Win7 has built-in drivers, it should find it and install the drivers, no "external" or "installation" media required.
• Now remove the disk (VDI) from the SCSI controller and add it to the SATA one.
• Boot the VM to make sure that everything is in order.
• Remove the SCSI controller altogether.

BACKUP FIRST!!!

Thank you. This seems to have resolved the issue.

[Socratis]

@morr_rva
Not really, it's a workaround to get you started. The ticket is still valid and it has to be addressed.

There are two workarounds it seems:

• Switch to SATA:AHCI.
• Don't use saved states... ;)

[Andy]

Debian 9 guest resuming from saved state

[Andy]

Same issue, different host/guest. Attaching logs in case it's any help with debugging.

[andronat]

Same issue for me when I'm using Vagrant under Host: macos 10.14.3 and guest: ubuntu 18.04.1.

➜  srg-website git:(master) ✗ vagrant up && vagrant ssh
Bringing machine 'default' up with 'virtualbox' provider...
==> default: Checking if box 'ubuntu/bionic64' version '20181114.1.0' is up to date...
==> default: Resuming suspended VM...
==> default: Booting VM...
There was an error while executing `VBoxManage`, a CLI used by Vagrant
for controlling VirtualBox. The command and stderr is shown below.

Command: ["startvm", "c3bdd69a-c569-4779-bdeb-82cb9a5054e3", "--type", "headless"]

Stderr: VBoxManage: error: Failed to load unit 'lsilogicscsi' (VERR_SSM_LOADED_TOO_LITTLE)
VBoxManage: error: Details: code NS_ERROR_FAILURE (0x80004005), component ConsoleWrap, interface IConsole

[mzb]

Same issue for me after suspend / restore - also using Vagrant: VirtualBox: 6.0.2 host: Windows 10 guest: Ubuntu 16.04

Failed to open a session for the virtual machine web3_default_1548545169711_21792.

Failed to load unit 'lsilogicscsi' (VERR_SSM_LOADED_TOO_LITTLE).

Result Code: E_FAIL (0x80004005) Component: ConsoleWrap Interface: IConsole {872da645-4a9b-1727-bee2-5585105b9eed}

[msys]

I'm also affected but both guest and host are Windows 7 64-bit (so no Linux is involved).

Also, the same problem occurs with SAS controller (the same error but with 'lsilogicsas' instead).

Relevant forum post: ​https://forums.virtualbox.org/viewtopic.php?f=6&t=91520

[Swen Thümmler]

Same problem here with SAS controller (Arch Linux host, Windows 7 guest). Restore does no longer work. VirtualBox 6.0.4 with matching guest additions.

[aeichner]

Confirmed, will be fixed in the next maintenance release. Thanks for the reports!

[ci-zephyurus]

Replying to aeichner:

Confirmed, will be fixed in the next maintenance release. Thanks for the reports!

I got hit with the same issues twice in a row. I am using VBox 6.0.4 r128413 for a few weeks no.q

I am using windows 64-bit host and a 64-bit linux Debian GNU/Linux guest. I was requsted to reboot the windows PC once to install the latest update of Norton Internet security and then, today, for Windows update. I got hit with the same error first while I performed Norton update, and I thought I would look into the matter when I forgot the issue and let it reboot to finish the windows update.

An early fix would be very much appreciated. I had a reason to choose SCSI, but I have forgotten exactly what. I am going to bite the bullet and switch to SATA AHCI for now.

[aeichner]

Please try the latest 6.0 testbuild from here. It should contain the fix.