[qubes-users] qrexec_timeout does not truly accept 3600

qvm-prefs vmName qrexec_timeout 3600
does not return an error message. When read, 3600 is returned.
However, the VM is forcibly stopped after 15 minutes.

1800 works
2700 works
3600 is not honored

How can i get by this so this one VM can do finish its upgrade before forcibly rebooted?

1 Like

Rob Townley:

qvm-prefs vmName qrexec_timeout 3600
does not return an error message. When read, 3600 is returned.
However, the VM is forcibly stopped after 15 minutes.

1800 works
2700 works
3600 is not honored

How can i get by this so this one VM can do finish its upgrade before
forcibly rebooted?

Think you can disable qrexec with "qvm-features [vmname] qrexec 0". Do this on the template. Set back to 1 when done.

1 Like

I really need this qube to run for over an hour without the hypervisor killing it. Short of buying a much faster brand new machine, how does one do that?

There's a confusion here - qrexec-timeout is the time to wait on boot
for the qrexec agent to be connected.
If the qube isn't booting up then you have no chance of performing an
upgrade. You should fix that problem first. BUT...
The fact that "the VM is forcibly stopped after 15 minutes" suggests to
me that you have the idleness monitor enabled, since this *does* have a
default shutdown time of 15 mins.
Can you check to see if you have the shutdown-idle service enabled this
qube?

1 Like

awokd, will try qvm-features --verbose vmName qrexec 0

Replying directly here in Google Groups because could not find this post in email.

Where does one find the idleness monitor setting?

[root@dom0 ~]# qvm-prefs --verbose vmName | egrep -i shut
shutdown_timeout D 60
[root@dom0 ~]# qvm-prefs --verbose vmName | egrep -i idle
[root@dom0 ~]#

You should be able to see if it is set with qvm-features, or
qvm-service.
It is called (in Debian) shutdown-idle service