IncusOS fails to bring up networking

Hi, I’m having problems after installing IncusOS 202511292320 today on an ASUS NUC 15 Pro (Intel Ultra 7 255H with Intel 2.5G igc NIC). The IncusOS networking fails to return any information

When I go to the web UI, and select the networking tab I get:

When I try to show network information via the incus cli commandincus admin os system network show the response is:

Error: open /sys/class/net/_pc84bd6d7eda7/speed: no such file or directory

The network resource block shows from incus admin os system resources show

network:
  cards:
  - driver: igc
    driver_version: 6.17.9-zabbly+
    firmware_version: 2023:889d
    numa_node: 0
    pci_address: "0000:56:00.0"
    ports:
    - address: 88:ae:dd:67:6d:65
      auto_negotiation: true
      id: _p88aedd676d65
      link_detected: true
      link_duplex: full
      link_speed: 2500
      port: 0
      port_type: twisted pair
      protocol: ethernet
      supported_modes:
      - 10baseT/Half
      - 10baseT/Full
      - 100baseT/Half
      - 100baseT/Full
      - 1000baseT/Full
      - 2500baseT/Full
      supported_ports:
      - twisted pair
      transceiver_type: internal
    product_id: 125c
    vendor_id: "8086"

I’d appreciate any help understanding why the networking fails to load and how best to resolve it, thanks.

Sounds like some kind of kernel issue as you’d expect that particular sysfs entry to exist.

Thanks! That PR looks like it matches the error I’m seeing. Once it’s available in an update, I’ll test again and confirm that it is resolved.

Hi again, I’ve updated the OS to the testing channel version 202512022133 that contains the PR merge and I can confirm that I’m now able to see the network info from commandincus admin os system network show along with the web ui networking tab.

But now I’m not able to edit the network config… :melting_face: I’m trying to edit the network config to include the roles tag instances to attach instances to my host network.

When I try to edit the config using , to add the role instances tag to the config from:

config:
  interfaces:
  - addresses:
    - dhcp4
    - slaac
    hwaddr: 88:ae:dd:67:6d:65
    name: enp86s0
    required_for_online: "no"   
  - addresses:
    - dhcp4
    - slaac
    hwaddr: c8:4b:d6:d7:ed:a7
    name: c84bd6d7eda7
    required_for_online: "no"
  time:
    timezone: UTC

to

config:
  interfaces:
  - addresses:
    - dhcp4
    - slaac
    hwaddr: 88:ae:dd:67:6d:65
    name: enp86s0
    required_for_online: "no"
    roles:
    - instances
  - addresses:
    - dhcp4
    - slaac
    hwaddr: c8:4b:d6:d7:ed:a7
    name: c84bd6d7eda7
    required_for_online: "no"
  time:
    timezone: UTC

I get the following error: Config parsing error: Put https://192.168.68.53:8443/os/1.0/system/network` read tcp 192.168.68.56:53416->192.168.68.53:8443: wsarecv: A connection attempt failed because the connected party did not properly respond after a period of time, or established connection failed because connected host has failed to respond. Press enter to open the editor again or ctrl+c to abort change`

In the incus os display, there is an error output message:

Config parsing error: timed out waiting for udev to rename interface(s)

When then abort with ctrl+c to abort the change, I lose access to incusOS entirely. Any commands I make as a follow up will all timeout

Error: Get "https://192.168.68.53:8443/1.0": Unable to connect to: 192.168.68.53:8443 ([dial tcp 192.168.68.53:8443: i/o timeout])

The ui is also no longer accessible. The incusOS display drops no longer displays the ip for enp86s0 as well in the networking configuration section section. This is only resolvable once I shutdown the machine and turn it back on which then successfully retrieves the network config minus my desired changes.

I’d appreciate any help on this. It could be that I was too hasty to update to a testing channel build rather than wait for stable but I didn’t think I had seen any commits that would have related to the network edit command for the version.

Try to only do dhcp4 and slaac on a single interface.

Doing DHCP on multiple interfaces will lead to undefined behavior where depending on which interface comes up first, traffic will head through it, potentially breaking connections.

We do it during setup as we have no idea what interface may be the correct one to use, but once you push your own config, it’s a good idea to avoid this.

Brilliant! All works now. I can edit and attach instances to my host network. Thanks for all the help on this :smile: