`incus cluster join` fails with "lookup unix.socket on 127.0.0.53:53: no such host"

I have two newly-installed IncusOS servers; I enabled clustering on the one, and tried to join the other one to the cluster with incus cluster join <cluster>: <node2>:, but got the error:

Error: Failed to join cluster: Failed to setup cluster trust: Failed to add server cert to cluster: Post "https://unix.socket:8443/1.0/certificates": lookup unix.socket on 127.0.0.53:53: no such host

The second joining node is a fresh install without apply_defaults, and the first clustered node is a day old with a couple test instances. They both have almost no configuration set:

incus config show <cluster>::

config:
  cluster.https_address: <node1 ip>
  core.https_address: :8443
  storage.backups_volume: local/backups
  storage.images_volume: local/images
  storage.logs_volume: local/logs 

incus config show <node2>::

config:
  cluster.https_address: <node2 ip>

Both servers and my local machine are running 7.3. I am aware 7.4 recently released, but from a quick look at the commit titles it seems nothing relevant changed, although I could easily be wrong.

I tried to work around this by using a preseed, shown below, applied with a factory reset, but got this error:

ERROR Failed to configure cluster: Failed to initialize member: Failed to initialize storage pools and networks: Failed to create storage pool "local": Loop backed pools aren't supported in IncusOS

factory reset seeds.incus in yaml format:

version: "1"
# apply_defaults: true
preseed:
  certificates:
    - <my admin certificate>
  config:
    cluster.https_address: <node2 ip>
  cluster:
    enabled: true
    server_address: <node2 ip>
    cluster_address: "<node1 ip>:8443" # Seems to default to :443
    server_name: <node2 new name>
    cluster_token: <cluster join token>
    cluster_certificate: |-
      <cluster cert PEM>

Incus seems really cool so far, and I’m looking forward to hopefully getting a working cluster soon!

Can you check in your ~/.config/incus/config.yml if you happen to have keepalive enabled on either or both of those remotes?