Failed to start Qubes DB agent after upgrade fedora 24 to 25 (r3.2)

[Pawel G]

I upgraded dom0 from fedora-24 to fedora-25 on Qubes r3.2.

After reboot, I can login, GUI starts, I'm able to open dom0 Terminal window etc. but Qubes Manager is not running, I can't start VMs:

"Failed to execute child process "qvm-run" (No such file or directory)."

"journaltcl | grep Fail" shows:

"Failed to start Qubes DB agent", and
"Failed to start Virtualzation daemon".

I tried to restart qubesd (not sure if it's related), but it says that:

"Failed to restart qubesd.service: Unit qubesd.service failed to load: No such file or directory".

I can't find any related thread to my problem (here and issues on github).

Any help would be highly appreciated as I'm smart but unfortunately not too bright. I have no idea how to diagnose this problem and how to make my Qubes works again.

Thanks!

Pawel

[awokd]

On Wed, January 10, 2018 5:13 pm, Pawel G wrote:
> I upgraded dom0 from fedora-24 to fedora-25 on Qubes r3.2.

Well there's the problem! :) Ordinarily, you should not upgrade dom0.
There's an article somewhere that explains why, but basically it's not
touched by the other VMs so there's no reason.

If you really want to do it, check out fpeitre's "3.3" fork on Github.

[Andrew David Wong]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512

On 2018-01-10 16:30, 'awokd' via qubes-users wrote:
> On Wed, January 10, 2018 5:13 pm, Pawel G wrote:
>> I upgraded dom0 from fedora-24 to fedora-25 on Qubes r3.2.
>
> Well there's the problem! :) Ordinarily, you should not upgrade
> dom0. There's an article somewhere that explains why, but basically
> it's not touched by the other VMs so there's no reason.
>

You might be thinking of this:

https://www.qubes-os.org/doc/supported-versions/#note-on-dom0-and-eol

> If you really want to do it, check out fpeitre's "3.3" fork on
> Github.
>

- --
Andrew David Wong (Axon)
Community Manager, Qubes OS
https://www.qubes-os.org

-----BEGIN PGP SIGNATURE-----
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=WXZw
-----END PGP SIGNATURE-----

[Pawel G]

On Wednesday, 10 January 2018 23:30:42 UTC+1, awokd wrote:
> On Wed, January 10, 2018 5:13 pm, Pawel G wrote:
> > I upgraded dom0 from fedora-24 to fedora-25 on Qubes r3.2.
>
> Well there's the problem! :) Ordinarily, you should not upgrade dom0.

Sorry for misunderstanding, I figured out what I did analysing my bash_history file (as I was sure that I haven't experimenting, just followed docummentation):

I upgraded dom0 (following "Installing and using Windows-based AppVMs" doc):

sudo qubes-dom0-update --enablerepo=qubes-dom0-current-testing qubes-windows-tools

and not long after that I upgraded all my VMs from f-24 to f-25.

I use my laptop for a few weeks puting it into hibernation mode. After restart - you know the story. I haven't found anything in bash_history what would damage qubes, aside from those updates.

Is it possible to reinstall base system and have my VMs and data not earased (like from ISO installer)?

Thanks.

[awokd]

On Thu, January 11, 2018 7:37 pm, Pawel G wrote:

> I upgraded dom0 (following "Installing and using Windows-based AppVMs"
> doc):
>
>
> sudo qubes-dom0-update --enablerepo=qubes-dom0-current-testing
> qubes-windows-tools
>
> and not long after that I upgraded all my VMs from f-24 to f-25.

OK, that's different, that should be fine. dom0 is still on f-24 then,
just your VMs are upgraded.

>
> I use my laptop for a few weeks puting it into hibernation mode. After
> restart - you know the story. I haven't found anything in bash_history
> what would damage qubes, aside from those updates.

I don't know what could have broken your QubesDB then, unless maybe you
ran just "sudo qubes-dom0-update --enablerepo=qubes-dom0-current-testing"
at some point? That would have upgraded Xen and everything except the VMs
to the testing repo, which is not always stable.

> Is it possible to reinstall base system and have my VMs and data not
> earased (like from ISO installer)?

If you back up the VMs first, you can reinstall then restore the VMs. In
3.2, go to Qubes VM Manager, then System menu, and choose Backup. Best if
you can do this to an external drive so the backups don't get wiped by
mistake on the install!

[Pawel G]

On Thursday, 11 January 2018 21:12:22 UTC+1, awokd wrote:

> If you back up the VMs first, you can reinstall then restore the VMs. In
> 3.2, go to Qubes VM Manager, then System menu, and choose Backup. Best if
> you can do this to an external drive so the backups don't get wiped by
> mistake on the install!

Is this the only way? No recovery mode? :) Qubes VM Manager is not running unfortunetly, so no go for me.

[awokd]

You can run qvm-backup from the command line. I thought there was a way to
start the GUI one too but am not finding it.

[pa...@kc.pl]

On Thursday, 11 January 2018 21:58:32 UTC+1, awokd wrote:

> You can run qvm-backup from the command line. I thought there was a way to
> start the GUI one too but am not finding it.

I have no qvm* commands available. Nor qubes-dom0-*.
Probably I can't also attach any external device at this point.

I have only qubesctl command and qubesdb-* commands available.

[Pawel G]

I found out some more information regarding qubes-db-dom0.service failure:

symlink /var/run/qub es/qubesdb.sock: No such file or directory
FATAL: server socket initialization failed
qubes-db-dom0.service: Main process exited, code=exited, status=1/FAILURE
Failed to start Qubes DB agent.

[awokd]

Sorry then, no idea how that would have happened. You can maybe copy out
the VMs' image files directly from /var/lib/qubes/appvms and mount them
after repairing your install to recover the contents.