Member discovery fails: discovery.talos.dev not reachable (1.9.1, Openstack)

[tbshrst]

Attempting to form a cluster consisting of 1 control plane and 1 worker node fails, because both machines cannot reach discovery.talos.dev for member discovery.

Both are running in Openstack in the same net with an external gateway configured for internet access. An addition, both nodes are configured to use a proxy.

Logs:
talosctl logs controller-runtime

ERROR hello failed {"component": "controller-runtime", "controller": "cluster.DiscoveryServiceController", "error": "rpc error: code = Unavailable desc = connection error: desc = "transport: Error while dialing: dial tcp 172.174.35.21:443: connect: no route to host"", "endpoint": "discovery.talos.dev:443"}
ERROR hello failed {"component": "controller-runtime", "controller": "cluster.DiscoveryServiceController", "error": "rpc error: code = Unavailable desc = connection error: desc = "transport: Error while dialing: dial tcp 172.174.35.21:443: i/o timeout"", "endpoint": "discovery.talos.dev:443"}
ERROR hello failed {"component": "controller-runtime", "controller": "cluster.DiscoveryServiceController", "error": "rpc error: code = Unavailable desc = connection error: desc = "transport: Error while dialing: dial tcp 172.174.35.21:443: i/o timeout"", "endpoint": "discovery.talos.dev:443"}
..

controlplane.yaml/worker.yaml:

machine:
	env:
    HTTP_PROXY: http://$IP
    HTTPS_PROXY: http://$IP
    http_proxy: http://$IP
    https_proxy: http://$IP

However, if I run the discovery service locally both machines can connect.
Could it be possible that the DiscoveryController ignores the proxy settings?

Environment
Talos version: v1.9.1
Kubernetes version: v1.32.0
Platform: Openstack

[smira]

No, Discovery Service respects proxy settings. These failures might have happened before you apply proxy settings, but totally not clear from your description.

Use talosctl get members to see if Discovery Service is working.

[chrxmvtik]

No, Discovery Service respects proxy settings. These failures might have happened before you apply proxy settings, but totally not clear from your description.

Use talosctl get members to see if Discovery Service is working.

We have the same issue, I can connect to discovery service via our corporate proxy from within container in the cluster, but Talos nodes can't (both are using the same proxy).

Seems like requests to discovery service are somehow not sent using proxy settings.

We have both lower and upper case environment variables set.

INB4: Each machine states, that there is just one machine in the cluster (itself), while we have multiple nodes.

Talos: 1.9.4
Kubernetes: 1.32.2

[smira]

yes, this seems to be a bug in this specific place around discovery service.