All VM guests now hang at transition to "Starting"

[eau sèche]

Greetings, and a request for help.

Shortly after updating the both the guest additions and the kernel, I noticed that my VMs were not running. The new kernel didn't start, so the older kernel is still used.

The host is Centos 7, and I have 2 VM guests. One guest runs Windows 7 while the other runs Ubuntu 64. Neither guest runs, but I've attempted most of my debugging with the Linux guest.

Please see the attachments for:

• - GUI screenshot
  - VBox.log
  - VBoxSVC.log
  - VBoxManage commands
  • - hostinfo
    - showvminfo
    - debugvm osinfo
    - debugvm info
    - debugvm dumpvmcore
    - list runningvms
    - list vms

The errors in the last 7 lines of "VBoxSVC.log" were not there in earlier logs, but appeared only after removing and adding the storage as a trial fix.
I've done my best to search for solutions, but don't find any that apply to the case here.
Experimental downgrades to VirtualBox and the guest additions have the same behaviour.

It appears that the VM doesn't actually start, per the "list runningvms". The "debugvm" commands seem to support this.

All files in the VM user's home directory structure are owned by the user "vbox".

I'm at a loss for how to proceed and would welcome suggestions.

[scottgus1]

Vbox.log dies off too quick perhaps to get anything good from it.

Showvminfo's die off at checking USB.

Going to guess that something really went wrong and you need to completely uninstall & purge Virtualbox and the Extension Pack, then reinstall from https://www.virtualbox.org/wiki/Linux_Downloads (not from your distro's repositories, those are forks we don't support here). I know you mentioned that you downgraded, try again with official Virtualbox. Also move the /home/{you}/.config/Virtualbox folder to your desktop and register your guests with the empty Virtualbox, see if they start.

[eau sèche]

Thanks for your response.

I was using the repo at virtualbox.org, but I erased the installation and installed the rpm directly after downloading from https://www.virtualbox.org/wiki/Linux_Downloads. The ../.config/Virtualbox subdirectory was removed.

After adding the virtual machine to the new installation, the behaviour was the same.

The attachments below contain the new log files, as well as logs of 'ps', 'gcore', 'gdb' with backtrace. All were obtained while the startup was hung.

Do these suggest anything, or are there other ways to investigate my problem?

[eau sèche]

Making some progress, even though a little confusing. Still in progress, but sequence leading to some success:

• -continuing from fresh installation per your suggesions
  - created new VM, attaching the existing disk image
  - same results (hang) when trying to start the new machine
  - killed process to unhang (as done many times before)
  - obtained a new copy of the same appliance from bitnami as an .ova file
  - restarted virtualbox
  - imported appliance
  - started new bitnami VM, which booted normally!
  - shutdown the new bitnami VM
  - tried starting the VM with my existing disk image, and it worked!!

I've got more work to do, but wanted to intercept any added help based on my previous post.

[eau sèche]

It took some unwinding to get back to all configurations before the debugging exercise began, but now both Windows and Ubuntu VMs are running with the original *.vbox configurations and virtual disks.

The fix was unsatisfying because the original cause is unknown. It seems that the actions that cleared the problem were one or both of:

• - importing a new virtual appliance
  - running the newly imported virtual appliance

Thank you to scottgus1 for your assistance.

[scottgus1]

Problems that disappear without solid known reasons certainly are unconfortable. Interesting that getting a new copy of the guest worked. Maybe the old one was bad, or a stray cosmic ray... Glad you're back running!

[eau sèche]

Stranger than a new guest running. It fixed the old guest!