tl;dr
incus CLI is still trying to use an IP my laptop doesn’t use anymore (used to be static, could have been setup locally at some point, then I moved to setting it up on my DHCP server and don’t remember if I left anything local that could cause all of this).
A simple command like:
incus conig get core.https_address
returns:
Error: Get "https://192.168.1.224:8443/1.0": EOF
That used to be my static local network IP, but now my local IPs are in 192.168.2.0/24, so there’s something from the old config messing things up.
Context
I tried incus with the interactive init some time ago. Created a few images and image pools, looked good for what I needed and I set it aside for a while.
Came back to it after moving home and changing subnet from 192.168.1.0/24 to 192.168.2.0/24 (and might change it again soon).
As I was trying the Cluster API infra provider for incus, I blindly followed the Getting started instructions, which include an incus admin init part.
Did a few rounds of apt uninstall, apt purge, manual zpool removals, manual brctl removals (I think I only needed this once), but to no avail.
incus version
Client version: 6.21
Server versions: unreachable
I haven’t found any reference to my old IP in an incus admin init –dump and I’m too new to incus to have any clue about where to look for text files containing the configuration (but have tried /opt/incus and /var/lib/incus too, although errors about large files might have gotten in the way of that).
Any pointers about where to look for the root cause would be very appreciated.
More logs/config dumps
systemctl status incus.service log lines:
Feb 12 18:02:38 myhostname dnsmasq-dhcp[132781]: IPv6 router advertisement enabled
Feb 12 18:02:38 myhostname dnsmasq-dhcp[132781]: DHCP, sockets bound exclusively to interface incusbr0
Feb 12 18:02:38 myhostname dnsmasq[132781]: using only locally-known addresses for incus
Feb 12 18:02:38 myhostname dnsmasq[132781]: reading /etc/resolv.conf
Feb 12 18:02:38 myhostname dnsmasq[132781]: using nameserver 127.0.0.53#53
Feb 12 18:02:38 myhostname dnsmasq[132781]: using only locally-known addresses for incus
Feb 12 18:02:38 myhostname dnsmasq[132781]: bad address at /etc/hosts line 3
Feb 12 18:02:38 myhostname dnsmasq[132781]: bad address at /etc/hosts line 4
Feb 12 18:02:38 myhostname dnsmasq[132781]: bad address at /etc/hosts line 6
Feb 12 18:02:38 myhostname dnsmasq[132781]: read /etc/hosts - 10 names
Those logs don’t match my machine’s /etc/hosts, not 10 names there, just the defaults and a couple alias lines for 127.0.0.1 to resolve hostnames of some docker tests I did.
Are the log lines referring to incus’s own /etc/hosts and resolf.conf? Where are hose in my filesystem?
incus admin init –dump
config: {}
networks:
- config:
ipv4.address: 10.109.81.1/24
ipv4.nat: "true"
ipv6.address: fd42:ccb0:147:5f4e::1/64
ipv6.nat: "true"
description: ""
name: incusbr0
type: bridge
project: default
storage_pools:
- config:
size: 30GiB
source: /var/lib/incus/disks/default.img
zfs.pool_name: default
description: ""
name: default
driver: zfs
storage_volumes: []
profiles:
- config: {}
description: Default Incus profile
devices:
eth0:
name: eth0
network: incusbr0
type: nic
root:
path: /
pool: default
type: disk
name: default
project: ""
projects:
- config:
features.images: "true"
features.networks: "true"
features.networks.zones: "true"
features.profiles: "true"
features.storage.buckets: "true"
features.storage.volumes: "true"
description: Default Incus project
name: default
certificates: []
cluster_groups: []