Odd behaviour with debian/12 and ubuntu/22.04 images on incus 6.7

Incus
1

Containers I try to create with the debian/12 and ubuntu/22.04 images are created without the default profile. Other images are adding the default profile just fine.

I noticed this when I was testing the instance creation scriptlet on incus 6.7.

pargo@bastion:~/scriptlet$ incus create images:debian/12 test-debian12
Creating test-debian12
Error: Failed instance creation: Failed instance placement scriptlet: Failed to run: fail: Núcleos de CPU não foram fixados.
pargo@bastion:~/scriptlet$ incus create images:debian/12/cloud test-debian12cloud
Creating test-debian12cloud
pargo@bastion:~/scriptlet$ incus create images:debian/11 test-debian11
Creating test-debian11
pargo@bastion:~/scriptlet$ incus create images:debian/11/cloud test-debian11cloud
Creating test-debian11cloud
pargo@bastion:~/scriptlet$ incus create images:ubuntu/22.04 test-ubuntu2204
Creating test-ubuntu2204
Error: Failed instance creation: Failed instance placement scriptlet: Failed to run: fail: Núcleos de CPU não foram fixados.
pargo@bastion:~/scriptlet$ incus create images:ubuntu/22.04/cloud test-ubuntu2204cloud
Creating test-ubuntu2204cloud
pargo@bastion:~/scriptlet$ incus create images:ubuntu/24.04 test-ubuntu2404
Creating test-ubuntu2404
pargo@bastion:~/scriptlet$ incus create images:ubuntu/24.04/cloud test-ubuntu2404cloud
Creating test-ubuntu2404cloud
pargo@bastion:~/scriptlet$

The failed error message is related to the scriptlet and could be ignored. The main issue I found through the scriptlet on the config request coming out empty as "config": {} and profiles comes out as "profiles": []. It consistently happens on those two images and not on the others.

Could this be an issue on the image server or could it be on my local cluster only?

2

Hi.

I tried this and was unable to re-create it. Maybe try deleting the local cached images and see if a refresh works? It’s not exactly a robust solution, but it could be a simple fix?

3

This is the exact error message. Why does it complain for CPU cores even if this is not about VMs?
You are launching a container, not a VM, anyway. Why?

Also, you are running the incus client with the Portuguese localization.
WHY is there no translated string for the message in the Portuguese localization?
Where does Incus pull this string translation from?

Not found in Incus/CLI — Portuguese (Brazil) @ Hosted Weblate
Not found neither in https://github.com/lxc/incus/blob/main/po/pt_BR.po

(note: the pt.po localization is empty, the pt_BR localization has translated 101 messages).

4

It’s not related to VMs. That error message was thrown by my placement scriptlet. It’s being launched in a project which requires all instances to be restricted to a few CPU cores. I’m the one throwing that error message since the instance configuration is not there (but it’s in the default profile).

5

This worked! Thank you very much!

Using incus image info I noticed there is a field for profiles. Maybe those were unset for some previously downloaded images and I was getting that problem.

$ incus image info 7fd26f41ea84
Fingerprint: 7fd26f41ea84264cfad062d6a66b097a3db2dd2b8ca206b7dabd211b7bddd78b
Size: 124.29MiB
Architecture: x86_64
Type: container
Public: no
Timestamps:
    Created: 2024/11/14 21:00 -03
    Uploaded: 2024/11/16 14:43 -03
    Expires: 1969/12/31 21:00 -03
    Last used: never
Properties:
    description: Ubuntu jammy amd64 (20241115_07:42)
    os: Ubuntu
    release: jammy
    serial: 20241115_07:42
    type: squashfs
    variant: default
    architecture: amd64
Aliases:
Cached: no
Auto update: enabled
Source:
    Server: https://images.linuxcontainers.org
    Protocol: simplestreams
    Alias: ubuntu/22.04
Profiles:
    - default
1 Like