This is because of too old template. Due to recent python3 version requirements for app-linux-pdf-converter, we had to create all the needed python38 dependencies because CentOS8 provides only python 3.6 packages (the one with python3- prefix). An extra COPR repository that I manage is for this purpose.
I notice these graphical issues when restoring Debian 10 VMs. How can I apply the same fixes like you have?
When trying to upgrade one of my Standalone Debian 10 R4.0 VMs. All I did was change the 4.0 to 4.1 for my qubes repo:
apt update; apt upgrade
Reading package listsā¦ Done
Building dependency tree
Reading state informationā¦ Done
Calculating upgradeā¦ Done
The following packages were automatically installed and are no longer required:
libblas3 libgfortran5 libimagequant0 liblapack3 libquadmath0 python-numpy
python-pil python-pkg-resources python3-numpy python3-pil
Use āsudo apt autoremoveā to remove them.
The following NEW packages will be installed:
libxengnttab1 libxentoolcore1 libxentoollog1 xserver-xorg-qubes-common
The following packages have been kept back:
qubes-core-agent qubes-core-agent-networking
The following packages will be upgraded:
libqrexec-utils2 libqubes-rpc-filecopy2 libqubesdb pulseaudio-qubes
python3-qubesdb qubes-core-agent-passwordless-root qubes-gui-agent
qubes-utils qubes-vm-dependencies qubesdb qubesdb-vm
xserver-xorg-input-qubes xserver-xorg-video-dummyqbs
13 upgraded, 4 newly installed, 0 to remove and 2 not upgraded.
Need to get 332 kB of archives.
After this operation, 121 kB of additional disk space will be used.
Do you want to continue? [Y/n]
Minus some noise here ā¦
ā¦
Installing new version of config file /etc/X11/Xsession.d/25xdg-qubes-settings
ā¦
Installing new version of config file /etc/X11/xorg-qubes.conf.template ā¦
Installing new version of config file /etc/xdg/autostart/qubes-icon-sender.desk
top ā¦
Setting up pulseaudio-qubes (4.1.11-1+deb10u1) ā¦
Processing triggers for libglib2.0-0:amd64 (2.58.3-2+deb10u2) ā¦
Processing triggers for libc-bin (2.28-10) ā¦
Setting up qubes-vm-dependencies (4.1.10-1+deb10u1) ā¦
/usr/lib/qubes/qrexec-client-vm: error while loading shared libraries: libqrexec-utils.so.2: cannot open shared object file: No such file or directory
ā
apt install libqrexec-utils2
Reading package listsā¦ Done
Building dependency tree
Reading state informationā¦ Done
libqrexec-utils2 is already the newest version (4.1.8-1+deb10u1).
libqrexec-utils2 set to manually installed.
apt list --upgradable
Listingā¦ Done
libvchan-xen/unknown 4.1.3-1+deb10u1 amd64 [upgradable from: 4.0.8-1+deb10u1]
qubes-core-agent-networking/unknown 4.1.16-1+deb10u1 amd64 [upgradable from: 4.
0.56-1+deb10u1]
qubes-core-agent/unknown 4.1.16-1+deb10u1 amd64 [upgradable from: 4.0.56-1+deb1
0u1]
apt upgrade
The following packages have been kept back:
qubes-core-agent qubes-core-agent-networking
0 upgraded, 0 newly installed, 0 to remove and 2 not upgraded.
And now after shutting down and relaunching the VM, it wonāt start.
You need to ensurethat you have latest gui-daemon in dom0 (current-testing) and to have latest gui-agent-linux in you VM (current-testing) too.
After fiddling with all .onion/https, 4.0/4.1 repo combos, I managed to upgrade those packages, but the last package threw an error. I cannot boot the VM again now.
Unpacking xen-utils-guest (4.14.0-7+deb10u1) ...
dpkg: error processing archive /var/cache/apt/archives/xen-utils-guest_4.14.0-7+deb10u1_amd64.deb (--unpack):
trying to overwrite '/lib/systemd/system/xendriverdomain.service', which is also in package xen-utils-common 2001:4.8.5-23+deb10u1
Errors were encountered while processing:
/var/cache/apt/archives/xen-utils-guest_4.14.0-7+deb10u1_amd64.deb
E: Sub-process /usr/bin/dpkg returned an error code (1)Updating the centos-8 template fails again, due to incompatibilities in python3, even if the repositories epel-8-python38 and epel-8-qubes are enabled via copr. The problem, which cannot be solved by specifying --allowerasing, --skip-broken and/or --nobest, seems to be a collision of several files (rst*) in /usr/bin between python38-docutils-0.16-3.el8.noarch and
python3-docutils-0.14-12.module_el8.3.0+562+e162826a.noarch.
CentOS 8 fixed for me.
Archlinux fixed for me, expect for pulseaudio, but that will come eventually.
Thank you to everyone who contributed to fixing the issue.
1 Like
Are there working ready arch linunx template and centos 8 for downloading over the community-templates-repo or I have to build it on my own?
And what exactly was the fix now that it is working? The latest-gui-agent?
If I need to build it I will try it the next time and update if it ist working for me:)
edit:
centos-8 is working for me now ,only pulse-audio in arch not .
but I am sure that this will be fixed soon and it doesnāt matter.
I build it on my own following these great descriptions here and the officiall qubes doc how to build it. But for people I found that archlinux minimal.rpm from Neowultran here:
https://neowutran.ovh/qubes/vm-archlinux/
If u download it and copy it to dom0 ,then run dnf install .rpm and after installing u can do a complete update & upgrade in the templatevm. Then it is working,too without the pulseaudio. That should be easier for people, who like to have an archlinux template and donāt know or have the time to run through the building steps.
much blesses
Thank you
I updated again at my other pc now and got the testing repo packages working renaming the qmemman.conf.rpmnew to .conf only and delete the old one. but the display in the archvm is still weird. what exactly is the fix for that? In a bionic vm it is the same problem. at my other pc everything is working ,but I donāt know what is missing now at the new one.
Ps. I cannot delete a post here. Otherwise I would replace my post before. sorry for that many posts. I dont really want that .
Fixed again and lucky for the weekend. And update and switch to a newer kernel in the qubes manager fixed it for me and now working again. Just my wired networking problem with new kernels I need to fix now and then I can run a backup of my lovely qubes system hehe;)
Ps: And always doing a reboot after doing such things affecting the system before trying it:)
Hey, I had to install centos-8 again at an other Pc and updating with --nobest --skip broken & --allowerasing helps to get the vm fully updated and audio working.
Only archlinux now has the audio problem. I have a read in an other topic that when I build the arch template on my own with qubes builder then audio is working? Is it worth to do those steps and can anybody confirm if it is working then? It does not really matter, but it is every time fine to have a complete working system:)
Thank you