Apt upgrade gets stuck on incus incus-client

The following packages will be upgraded:
incus incus-client
2 upgraded, 2 newly installed, 0 to remove and 0 not upgraded.
Need to get 85.4 MB of archives.
dpkg: warning: unable to delete old directory ‘/var/log/incus’: Directory not empty
dpkg: warning: unable to delete old directory ‘/var/lib/incus’: Directory not empty
Unpacking incus-client (0.5-202401260134-ubuntu20.04) over (0.4-202312232117-ubuntu20.04) …
Selecting previously unselected package incus-base.
Preparing to unpack …/incus-base_0.5-202401260134-ubuntu20.04_amd64.deb …
Unpacking incus-base (0.5-202401260134-ubuntu20.04) …
Setting up nftables (0.9.3-2) …
Setting up incus-client (0.5-202401260134-ubuntu20.04) …
Setting up incus-base (0.5-202401260134-ubuntu20.04) …
Progress: [ 76%] [####################################################################################…]
Progress: [ 76%] [####################################################################################…]
[1]+ Stopped apt
upgrade#############################################################…]

And then nothing works on upgraded nodes.
I have to manually kill it
Tried redoing apt upgrade, always gets stuck there.

Cluster list shows servers fully operational but containers are not functioning on those servers.

And you made sure to run it on all nodes at the same time?

This is a version update from 0.4 to 0.5, so the upgrade will remain stuck until all servers in the cluster reach the same point, at which point they’ll all unblock.

Do I need to reboot servers afterward for it to work. I have two more computers to upgrade and dont want to reboot all the servers now.

Nope, no need to reboot anything, just need to make sure that they’re all on the same version for the update to complete.

It worked, they recovered themselves after the last one was updated.
Anyway, it would be nice if Incus update was somehow separate than regular upgrade. It makes me nervous when 8 servers get stuck on upgrade :wink:
Also some other ideas/question.

  1. How can I disable ipv6 from present containers, from them ever showing up again.
  2. It would be nice if Containers had a description field like Servers do

But definitely working better than LXD so far

You probably just want to set ipv6.address=none on your Incus network.

The column exists, it’s just not shown by default, you can see it with incus list -cns46td for example.

Where do I set it? You probably just want to set ipv6.address=none on your Incus network.

incus network set NAME ipv6.address=none

root@RR5:/home/rr2020# @ network list
±---------±---------±--------±----------------±--------------------------±------------±--------±--------+
| NAME | TYPE | MANAGED | IPV4 | IPV6 | DESCRIPTION | USED BY | STATE |
±---------±---------±--------±----------------±--------------------------±------------±--------±--------+
| 10.0.0.0 | bridge | YES | 10.119.230.1/24 | fd42:266e:d72f:8869::1/64 | | 0 | CREATED |
±---------±---------±--------±----------------±--------------------------±------------±--------±--------+
| enp1s0f0 | physical | NO | | | | 0 | |
±---------±---------±--------±----------------±--------------------------±------------±--------±--------+
| enp1s0f1 | physical | NO | | | | 0 | |
±---------±---------±--------±----------------±--------------------------±------------±--------±--------+
| incusbr0 | bridge | YES | 10.126.149.1/24 | fd42:ecf:4f02:b679::1/64 | | 17 | CREATED |
±---------±---------±--------±----------------±--------------------------±------------±--------±--------+
root@RR5:/home/rr2020# incus network incusbr0 ipv6.address=none
Error: unknown command “incusbr0” for “incus network”
root@RR5:/home/rr2020# incus network set incusbr0 ipv6.address=none
Error: failed to notify peer 192.168.1.50:8443: The DNS and DHCP service exited prematurely: exit status 2 (“dnsmasq: failed to create listening socket for 10.126.149.1: Address already in use”)
root@RR5:/home/rr2020# @ network list
±---------±---------±--------±----------------±--------------------------±------------±--------±--------+
| NAME | TYPE | MANAGED | IPV4 | IPV6 | DESCRIPTION | USED BY | STATE |
±---------±---------±--------±----------------±--------------------------±------------±--------±--------+
| 10.0.0.0 | bridge | YES | 10.119.230.1/24 | fd42:266e:d72f:8869::1/64 | | 0 | CREATED |
±---------±---------±--------±----------------±--------------------------±------------±--------±--------+
| enp1s0f0 | physical | NO | | | | 0 | |
±---------±---------±--------±----------------±--------------------------±------------±--------±--------+
| enp1s0f1 | physical | NO | | | | 0 | |
±---------±---------±--------±----------------±--------------------------±------------±--------±--------+
| incusbr0 | bridge | YES | 10.126.149.1/24 | fd42:ecf:4f02:b679::1/64 | | 17 | CREATED |
±---------±---------±--------±----------------±--------------------------±------------±--------±--------+

Can you show netstat -lnp on the machine with IP 192.168.1.50?

Use apt-mark hold incus to avoid the upgrade. Change hold to unhold afterwards to upgrade it.