Etcd don't start after bootstrap

[Minivolk02]

Bug Report

Etcd don't start after bootstrap

Description

Etcd don't start after bootstrap

Logs

[talos] serviceetcd: Condition failed: context canceled
user: warning: [2024-05-09T09:20:09.404547126Z]: [talos] serviceetcd: Bootstrap requested
user: warning: [2024-05-09T09:20:09.405089126Z]: [talos] serviceetcd: Starting service
user: warning: [2024-05-09T09:20:09.405594126Z]: [talos] serviceetcd: Waiting for service "cri" to be
"up", time sync, network, etcd spec
user: warning: [2024-05-09T09:20:09.406492126Z]: [talos] task startAllServices (1/1): done, 37.871156217s
user: warning: [2024-05-09T09:20:09.406984126Z]: [talos] phase startEverything (16/16): done, 37.872123767s
user: warning: [2024-05-09T09:20:09.407507126Z]: [talos] boot sequence: done: 38.463758891s
user: warning: [2024-05-09T09:20:10.407322126Z]: [talos] serviceetcd: Waiting for etcd spec

Environment

• Talos version:
  Client:
  Tag: v1.7.0
  SHA: 70fb41f
  Built:
  Go version: go1.22.2
  OS/Arch: linux/amd64
  Server:
  NODE: 10.224.238.2
  Tag: v1.7.0
  SHA: 70fb41f
  Built:
  Go version: go1.22.2
  OS/Arch: linux/amd64
  Enabled: RBAC
• Kubernetes version:
  Client Version: v1.30.0
  Kustomize Version: v5.0.4-0.20230601165947-6ce0bf390ce3
• Platform:
  Vagrant and Libvirt

[smira]

Please submit your issues with some logs, or preferably talosctl support bundle.

If I would guess, etcd spec is not built as there are no IPs to assign the etcd cluster to.

[Minivolk02]

support.zip
Here it is

[smira]

You specifically configure etcd not to advertise the only address you have on the node:

metadata:
    namespace: etcd
    type: EtcdConfigs.etcd.talos.dev
    id: etcd
    version: 1
    owner: etcd.ConfigController
    phase: running
    created: 2024-05-15T08:11:54Z
    updated: 2024-05-15T08:11:54Z
spec:
    advertiseExcludeSubnets:
        - 10.224.238.2
    listenExcludeSubnets: []
    image: gcr.io/etcd-development/etcd:v3.5.13
    extraArgs: {}

So etcd can't start, as it needs an address to advertise to other peers, and there is none available.

If you hit an issue, it's always a good idea to revert machine config changes and try with plain defaults. If defaults work, the problem might be your config change.