Cannot connect to qrexec agent for 60 seconds - on starting app qube

I recently created a new app qube now for the first time using the in-place upgrade on debian-12. My app qube fails to start with:

Cannot connect to qrexec agent for 60 seconds, see /var/log/xen/console/guest-A-python.log for details

The corresponding log file ends on a lot of lines looking like this:

.[K[.[0;31m*.[0;1;31m*.[0m.[0;31m*   .[0m] A start job is running for Adjust r…t filesystem size (43s / no limit)
.[K[ .[0;31m*.[0;1;31m*.[0m.[0;31m*  .[0m] A start job is running for Adjust r…t filesystem size (44s / no limit)
.[K[  .[0;31m*.[0;1;31m*.[0m.[0;31m* .[0m] A start job is running for Adjust r…t filesystem size (44s / no limit)
.[K[   .[0;31m*.[0;1;31m*.[0m.[0;31m*.[0m] A start job is running for Adjust r…t filesystem size (45s / no limit)
.[K[    .[0;31m*.[0;1;31m*.[0m] A start job is running for Adjust r…t filesystem size (45s / no limit)
.[K[     .[0;31m*.[0m] A start job is running for Adjust r…t filesystem size (46s / no limit)
.[K[    .[0;31m*.[0;1;31m*.[0m] A start job is running for Adjust r…t filesystem size (46s / no limit)
.[K[   .[0;31m*.[0;1;31m*.[0m.[0;31m*.[0m] A start job is running for Adjust r…t filesystem size (47s / no limit)
.[K[  .[0;31m*.[0;1;31m*.[0m.[0;31m* .[0m] A start job is running for Adjust r…t filesystem size (47s / no limit)
.[K[ .[0;31m*.[0;1;31m*.[0m.[0;31m*  .[0m] A start job is running for Adjust r…t filesystem size (48s / no limit)
.[K[.[0;31m*.[0;1;31m*.[0m.[0;31m*   .[0m] A start job is running for Adjust r…t filesystem size (48s / no limit)
.[K[.[0;1;31m*.[0m.[0;31m*    .[0m] A start job is running for Adjust r…t filesystem size (49s / no limit)
.[K[.[0m.[0;31m*     .[0m] A start job is running for Adjust r…t filesystem size (49s / no limit)
.[K[.[0;1;31m*.[0m.[0;31m*    .[0m] A start job is running for Adjust r…t filesystem size (50s / no limit)
.[K[.[0;31m*.[0;1;31m*.[0m.[0;31m*   .[0m] A start job is running for Adjust r…t filesystem size (50s / no limit)
.[K[ .[0;31m*.[0;1;31m*.[0m.[0;31m*  .[0m] A start job is running for Adjust r…t filesystem size (51s / no limit)
.[K[  .[0;31m*.[0;1;31m*.[0m.[0;31m* .[0m] A start job is running for Adjust r…t filesystem size (51s / no limit)
.[K[   .[0;31m*.[0;1;31m*.[0m.[0;31m*.[0m] A start job is running for Adjust r…t filesystem size (52s / no limit)

I tried increasing the private storage max size (the other one is grayed out) from 2 to 5GB but the error persists.

I had a similar thing happening a few days ago after resizing.
I increased the timeout to 300 seconds and after a while the qube started.

Try this in dom0:

qvm-prefs <qube name> qrexec_timeout 300

Another way is to tell systemd not to start the resize service using the kernelopts option, and to do the resizing manually after:

qvm-prefs <qube name> kernelopts systemd.mask=qubes-rootfs-resize.service

Also, if you have not already done so, make sure that the template is turned off.

1 Like

I just got bitten by this. I increased the debian-12-xfce template’s System storage max size from 42 GB to 50 GB. After that, the sys-firewall had the same problem logs as in the original post of this thread.

I did as well, using the command snippet you’ve provided. But now, logs tell me that the sys-net is taing 3 minutes to start. This is quite disturbing, the old startup speed of QubesOS is no more.


Is this bug something expected? Isn’t there a permanent solution to this, that doesn’t involve changing kernelopts stuff (which I don’t know anything about and makes me nervous to fiddle with)?

this doesn’t affect the problem, it persists even during the debian-12-xfce template is turned off (which it is already, most of the time).

To apply the change in the debian-12-xfce system storage size you need to start the debian-12-xfce template, wait for it to resize the system storage, shutdown debian-12-xfce template and then try to start app qubes based on this template.

1 Like

I started debian-12-xfce. It took it one or two seconds to finish booting up (compared to the 3 minute duration I reported above for sys-whonix. Now I will shutdown debian-12-xfce and shutdown my qubes and restart sys-firewall to see if it still takes 3 minutes to finish booting itself up.

If your sys-firewall/sys-net are disposables and you’ve increased the private storage for their disposable template then you need to start/stop the disposable template as well.

1 Like

they are disposables, yes. I increased the debian-12-xfce template’s system storage max size.


Anyways, interestingly, shutting down all my qubes one by one (including the sys-* ones), and restarting them back up, I see that sys-firewall is back again at 1 or 2 seconds boot up time! (instead of the 3 minutes) That’s great.

Hello everyone,

At the moment my TemplatVM is no longer opening, this is displayed on my screen.

has failed to start: cannot connect to qrexec agent for 60 seconds, see /var/log/xen/console/guest-debian-12-xfce .log for details

qvm-prefs <qube name> qrexec_timeout 300

Then open a console as the qube boots to see what the issue might be