Host doesn't seem to be checking for updates anymore?

IncusOS
1

One of my hosts got hit by the netlogd issue a couple weeks back, so I used the RECOVERY_DATA method to apply the update that fixed it. However that host doesn’t seem to be seeing any updates since? Is there some flag I need to clear for it to use the images.linuxcontainers.org source again after a RECOVERY_DATA update? My other two hosts both have 202609262332 staged, but I’m hesitent to reboot them incase they got pulled for some reason, and that’s why my one host isn’t seeing the update.

2

@Amanda what happens if you run incus admin os system update check and then look at incus admin os debug log -u incus-osd -n 30?

@gibmat FYI

3

Nothing gets printed in the logs, last message is still from yesterday’s “System is Ready”

4

What does incus admin os system provider show report? When performing an update from a RECOVERY_DATA source, we temporarily set the provider to “debug”, but then it should be automatically reverted. If that didn’t happen for you, that would explain the lack of updates on that system.

Switching the provider back to the default “images” one is simple. Just run incus admin os system provider edit with the following configuration:

config:
  name: images
  config:
    server_url: https://images.linuxcontainers.org/os
5

That seems to be it:

config:
  name: debug
state:
  registered: false
6

Hrm, though 24h later, it still doesn’t see any updates. I did the manual update again before I posted the original post, manually updating it to 202609222152, but still no updates are being suggested after fixing the provider.

7

Does incus admin os system provider show now properly report the provider as “images”?

If so, does manually running incus admin os system update check report any updates?

8

incus admin os system provider show says the following, and checking for updates still finds none, and doesn’t log anything

config:
  config:
    server_url: https://images.linuxcontainers.org/os
  name: images
state:
  registered: false

For completeness, after incus admin os system update check this is what incus admin os system update show says:

config:
  auto_reboot: false
  channel: stable
  check_frequency: 24h
state:
  last_check: "2026-10-01T06:43:06.846719934-04:00"
  needs_reboot: false
  status: Update check completed
9

Hrm. I’ve just checked incus admin os provider show on my other two nodes, both of which have been receiving updates and they both say:


config:
  name: images
state:
  registered: false
10

It’s odd that your system isn’t finding any updates. You could try removing the server_url configuration option, as the code will default to that value if no server URL is specified.

And since there’s no error reported in the update state, I’d assume the system is able to fetch the update metadata, so I’m not quite sure what’s going on here.

The output of the incus-osd logs might be useful: incus admin os debug log -b 0 -u incus-osd -n 300

11 
[2026/09/27 15:46:31 EDT] systemd: Started incus-osd.service - IncusOS - management daemon.
[2026/09/27 15:46:32 EDT] incus-osd: 2026-09-27 15:46:32 INFO System is starting up machine-id=65366f9d374b4d47895209f1045b9e2b mode=production version=202609222152
[2026/09/27 15:46:32 EDT] incus-osd: 2026-09-27 15:46:32 INFO Bringing up the network
[2026/09/27 15:47:04 EDT] incus-osd: 2026-09-27 15:47:04 WARN systemd-timesyncd failed to perform NTP synchronization, system time may be incorrect
[2026/09/27 15:47:06 EDT] incus-osd: 2026-09-27 15:47:06 INFO Starting service name=ceph
[2026/09/27 15:47:06 EDT] incus-osd: 2026-09-27 15:47:06 INFO Starting service name=tailscale
[2026/09/27 15:47:06 EDT] incus-osd: 2026-09-27 15:47:06 INFO Bringing up the local storage
[2026/09/27 15:47:07 EDT] incus-osd: 2026-09-27 15:47:07 INFO Starting application name=incus-ceph version=20.2.2 [202609222152]
[2026/09/27 15:47:07 EDT] incus-osd: 2026-09-27 15:47:07 INFO Starting application name=incus version=7.4 [202609222152]
[2026/09/27 15:47:21 EDT] incus-osd: 2026-09-27 15:47:21 INFO Application TLS certificate fingerprint fingerprint=119fcd0fe4ac103a3962375ed13b9a723743b1dee25bd420937856a51782a8ce name=incus
[2026/09/27 15:47:21 EDT] incus-osd: 2026-09-27 15:47:21 INFO System is ready version=202609222152

Nothing is showing up in the screen on the server as well when I try and do a incus admin os update check

12

@gibmat does the update checker get initialized with a particular provider maybe? that would explain what we’re seeing here if it doesn’t pull it fresh from state every time.

13

Hrm, that may have been it. I triggered a reboot, and it’s suddenly applying the latest update.

14

Hmm, yeah I guess we had a cached provider that was still looking at the “debug” configuration. The recent update refactoring work should help prevent that in the future, and I’ve also got PR up to fix the initial issue when applying recovery updates: Improve recovery logic when system reboots by gibmat · Pull Request #1403 · lxc/incus-os · GitHub.