csi-node fails to start with kube::Client hostname mismatch when contacting the k8s API server

[michaelbeaumont]

Describe the bug
csi-node fails to start:

["csi-node", "--csi-socket=/csi/csi.sock", "--node-name=k8rn-cp-0", "--rest-endpoint=http://openebs-openebs-api-rest:8081", "--enable-registration", "--grpc-endpoint=fd7a:115c:a1e0::ef01:c11:10199", "--nvme-io-timeout=110s10s", "--nvme-core-io-timeout=110s10s", "--nvme-ctrl-loss-tmo=1980", "--nvme-nr-io-queues=2", "--node-selector=openebs.io/csi-node=mayastor", "--fmt-style=pretty", "--ansi-colors=true"]
  2024-08-29T15:48:45.875422Z ERROR kube_client::client::builder: failed with error error trying to connect: error:0A000086:SSL routines:tls_post_process_server_certificate:certificate verify failed:ssl/statem/statem_clnt.c:1889:: hostname mismatch
    at /build/cargo-vendor-dir/kube-client-0.85.0/src/client/builder.rs:154

  2024-08-29T15:48:45.875504Z ERROR csi_node: Terminated with error, error: Failed to patch node: k8rn-cp-0 with patch: {"apiVersion":"v1","kind":"Node","metadata":{"labels":{"openebs.io/csi-node.nvme-ana":"true"}}}. HyperError: error trying to connect: error:0A000086:SSL routines:tls_post_process_server_certificate:certificate verify failed:ssl/statem/statem_clnt.c:1889:: hostname mismatch
    at control-plane/csi-driver/src/bin/node/main.rs:51

To Reproduce

I'm running a dual stack cluster and that's what seems to trigger this behavior for others: kube-rs/kube#991

The problem has already been solved with the rustls backend but the current version v0.85.0 uses openssl by default so I propose bumping kube to at least when rustls became the default (v0.87.2).

Expected behavior

Communication with the API server to work

OS info (please complete the following information):

• Distro: Talos
• 6.6.33
• openebs 4.1