!VhzbGHamdfMiGxpXyg:robins.wtf

NixOS Incus and LXC

52 Members
lxc, lxd, incus discussions related to NixOS19 Servers

Load older messages

SenderMessageTime
12 May 2026
@bart:bartoostveen.nlBart joined the room.18:09:16
16 May 2026
@tom:dragar.deTom

When rebuilding the unstable nixos incus image there is currently the follwing failure:

error: The option `programs.command-not-found.enable' has conflicting definition values:
       - In `/nix/store/wy93nw7bgjiy88dvpxfsj3wq8wgmzl35-nixos/nixos/nixos/modules/profiles/minimal.nix': false
       - In `/nix/store/wy93nw7bgjiy88dvpxfsj3wq8wgmzl35-nixos/nixos/nixos/modules/programs/command-not-found/command-not-found.nix': true
       Use `lib.mkForce value` or `lib.mkDefault value` to change the priority on any of these definitions.
15:17:11
@tom:dragar.deTomThere is the following PR https://github.com/NixOS/nixpkgs/pull/51278515:17:57
@tom:dragar.deTom *

When rebuilding the unstable nixos incus image with the default config there is currently the follwing failure:

error: The option `programs.command-not-found.enable' has conflicting definition values:
       - In `/nix/store/wy93nw7bgjiy88dvpxfsj3wq8wgmzl35-nixos/nixos/nixos/modules/profiles/minimal.nix': false
       - In `/nix/store/wy93nw7bgjiy88dvpxfsj3wq8wgmzl35-nixos/nixos/nixos/modules/programs/command-not-found/command-not-found.nix': true
       Use `lib.mkForce value` or `lib.mkDefault value` to change the priority on any of these definitions.
15:19:32
18 May 2026
@adam:robins.wtfadamcstephensnow that nspawn tests are first class, i'm tempted to spike out switching the incus tests to them. especially for aarch64 that could be a huge win (since aarch64 nested virtualization is a shit show)13:52:06
@hexa:lossy.networkhexathat may or may not work given the limitations nspawn brings13:58:48
@hexa:lossy.networkhexain most cases you just replace nodes with containers though13:58:59
@adam:robins.wtfadamcstephensi'm not too familiar with nspawn, any limitations come to mind that you think could be a problem?14:40:12
@hexa:lossy.networkhexa no kernel specific things, no setuid (no switching), no gui stuff, no container<->vm network interaction 14:46:18
@adam:robins.wtfadamcstephenskernel could be a problem, yeah15:22:11
25 May 2026
@tom:dragar.deTom(https://github.com/lxc/lxc-ci/pull/992)16:27:11
@tom:dragar.deTom

adamcstephens: looking at the job i'm wondering wheter the logic should be changed to a if [ "$RELEASE" = "nixos-25.11" ]; then as not do the repackaging for 26.05 and onward?

(i'm currently assuming that this wasn't done for 25.11 because that would've required changes? It's not explained anywhere as far as i can tell)

https://github.com/lxc/lxc-ci/blob/c76d66f24ec00222932431bb922e338d6ca20b01/jenkins/jobs/image-nixos.yaml#L44-L58

16:53:14
@tom:dragar.deTom *

adamcstephens: looking at the job i'm wondering wheter the logic should be changed to a if [ "$RELEASE" = "release-25.11" ]; then as not do the repackaging for 26.05 and onward?

(i'm currently assuming that this wasn't done for 25.11 because that would've required changes? It's not explained anywhere as far as i can tell)

https://github.com/lxc/lxc-ci/blob/c76d66f24ec00222932431bb922e338d6ca20b01/jenkins/jobs/image-nixos.yaml#L44-L58

16:55:27
@tom:dragar.deTom *

adamcstephens: looking at the job i'm wondering wheter the logic should be changed to a if [ "$RELEASE" = "25.11" ]; then as not do the repackaging for 26.05 and onward?

(i'm currently assuming that this wasn't done for 25.11 because that would've required changes? It's not explained anywhere as far as i can tell)

https://github.com/lxc/lxc-ci/blob/c76d66f24ec00222932431bb922e338d6ca20b01/jenkins/jobs/image-nixos.yaml#L44-L58

16:55:49
@adam:robins.wtfadamcstephensit was done for unstable only as an initial test of using purely our images. it's been a success so yes, setting != 25.11 sounds right and then we can remove completely the distrobuilder stuff when retiring it18:23:54
@adam:robins.wtfadamcstephensweird that there's no PR associated with this commit https://github.com/lxc/lxc-ci/commit/999a75491447910bcc6ea20be4928cc584600bb718:33:56
@adam:robins.wtfadamcstephensi don't have push access to that repo...18:34:06
@adam:robins.wtfadamcstephenshttps://github.com/lxc/lxc-ci/pull/94018:35:55
@tom:dragar.deTomyeah, noticed that github weirdness to19:38:54
@tom:dragar.deTom* yeah, noticed that github weirdness too19:39:01
@tom:dragar.deTomhttps://github.com/lxc/lxc-ci/pull/99419:39:30
26 May 2026
@adam:robins.wtfadamcstephensi'm wondering if we should mark 25.11 incus LTS as insecure once 26.05 is officially released. still seems to be no movement on security fixes12:06:03
@tom:dragar.deTom7.1.0 is due in around two days (28.05. https://github.com/lxc/incus/milestone/36) that probably also means 7.0.1 is gonna happen then and i kinda suspect 6.0.7 is also going to happen around that same time13:27:38
@tom:dragar.deTomwe'll see but it really isn't ideal13:29:41
@tom:dragar.deTom* we'll see, but it really isn't ideal13:30:01
@tom:dragar.deTom* we'll see, and it definitly isn't ideal13:30:59
@tom:dragar.deTombut especially since we have 7.0.0+15 on 25.11 with the incus package i also wouldn't feel too badly with just marking the lts package as insecure on 25.11. We then just shouldn't backport 7.1.0 to 25.1113:34:00
@tom:dragar.deTom* but especially since we have 7.0.0+15 on 25.11 with the incus package i also wouldn't feel too bad about just marking the lts package as insecure on 25.11. We then just shouldn't backport 7.1.0 to 25.1113:36:19
@tom:dragar.deTombtw. there is now a lts-7.0 branch for the zabbly package https://github.com/zabbly/incus/tree/lts-7.013:37:42
27 May 2026
@dag0bertz:matrix.orgDieselgert Baghetto joined the room.05:40:07

Show newer messages

Back to Room ListRoom Version: 10