Issue with qubes update

User Support General
1

Hello,

For few days I can’t make updates anymore
Launch updater stay empty

update1
update12048×1536 417 KB

update2
update22048×1536 191 KB

someone have a clue how to fix this issue ?

Thanks

2

Could you try with the command-line?

2 Likes
3

sudo qubes-dom0-update work fine
and sudo qubes-vm-update started but i have (error) : 0% for all of them

4

Please provide the error message(s).

1 Like
5

Maybe we’re having a common cause for sudden update issues. As my GUI updater crashed before I did a clean install, I used the CLI just to be sure.

dom0 had this error:

>>> Finished %post scriptlet: qubes-gui-daemon-selinux-0:4.3.15-1.fc41.x86_64                                                                                                              
>>> Scriptlet output:                                                                                                                                                                      
>>> Failed to resolve typeattributeset statement at /var/lib/selinux/targeted/tmp/modules/200/qubes-gui-daemon/cil:5                                                                       
>>> Failed to resolve AST                                                                                                                                                                  
>>> semodule:  Failed!                                                                                                                                                                     
>>>

Related:

I proceeded with the templates hoping it’s not too serious, some error’d relatively quick for no apparent reason (~ 20%), one took a long time to error (at ~ 52%) and only plain debian-13-xfce succeeded. It doesn’t display any errors. The others are Kicksecure and Whonix based templates.

However, in my Qubes i3 bar at the bottom of the screen I see “qubesd connection failed” next to the usual information. Don’t know what to make of it and if my setup is screwed now, which would be very, very unfortunate.

Two things that come into my mind are the dom0 update error and Kicksecure/Whonix introducing new meta packages (CLI vs. Server, I think previously everything was server).

Any ideas?

6

no more than that

3
32048×1536 208 KB

7

What’s the exit code if you wait it out?

Does --log DEBUG / --show-output help getting more info?

8

this bug happen few days ago as i already told and i guess was after one dom0 update but i cant be 100% sure about that

i do not have more info with --show-output

9

I just executed it again with --show-output, now some more updates were successful. Others weren’t.

I was flooded with W:Tried to start delayed item tor+http://deb.w5j6stm77zs6652pgsij4awcjeel3eco7kvipheu6mtr623eyyehj4yd.onion trixie/main amd64 open-link-confirmation all 3:8.2-1, but failed so often I can’t scroll above it anymore.

The error is followed by this:

whonix-gateway-18:out: /rw: 121.7 MiB (127639552 bytes) trimmed on /dev/xvdb
whonix-gateway-18:out: /: 361.2 MiB (378744832 bytes) trimmed on /dev/mapper/dmroot
whonix-gateway-18:out: /usr/sbin/policy-rc.d returned 101, not running 'restart dev-xvdc1-swap.service home.mount qubes-bind-dirs.service qubes-early-vm-config.service qubes-misc-post.service qubes-mount-dirs.service qubes-rootfs-resize.service qubes-sync-time.service qubes-sync-time.timer qubes-sysinit.service qubes-update-check.service qubes-update-check.timer qubes-updates-proxy-forwarder.socket usr-local.mount'
whonix-gateway-18:out: Setting up tor-ctrl (3:6.4-1) ...
whonix-gateway-18:out: Setting up libgd3:amd64 (2.3.3-13) ...
whonix-gateway-18:out: Setting up initializer-dist (3:8.5-1) ...
whonix-gateway-18:out: Setting up whonix-gateway-general-cli (3:26.9-1) ...
whonix-gateway-18:out: Setting up sdwdate-gui (1:13.8-1) ...
whonix-gateway-18:out: Installing new version of config file /etc/qubes-rpc/sdwdate-gui.ConnectCheck ...
whonix-gateway-18:out: 
whonix-gateway-18:err:

Executing it another time, it says that no updates are available. Very odd.

Edit: What I ended up doing is reboot the PC to see if it still works at all, which was the case, rename the templates to [name]-old, restore the clean templates from right after a fresh install about three weeks ago, switch the VM templates to back to their original name from the restore, delete the -old versions and run the update CLI command again with --max-concurrency 1 to do them as gentle as possible. All template updates worked perfectly.

Maybe it was a hiccup because of the dom0 upgrade issue. Now it seems to be good for me for now.
Always create backups.

The tool needs to finish execution before it displays any errors. Did it finish or did you stop it?

Also run echo $? after the execution to get the exit code as it’s supposed to indicate why it failed.

If your updates fail at 0%, I would check the UpdateVM. Have you properly configured it in your settings? Does it have confirmed working internet connectivity?
If so, I would duplicate one of the templates and try running the updates from within the template manually and see if it works or not to further debug the issue.

1 Like