utils/etcd.go is not using the v1beta4 EtcdAPICall timeout for etcd API calls

[neolit123]

we added the new constant EtcdAPITimeout and it's used in etcd.go:
https://github.com/kubernetes/kubernetes/blob/release-1.30/cmd/kubeadm/app/util/etcd/etcd.go#L252
but instead, it should be using the configurable Timeouts.EtcdAPICall from v1beta4.
with the singleton pattern kubeadmapi.GetActiveTimeouts().EtcdAPICall.Duration

i'm marking this as a bug since this is not intentional and the original exponential backoff was more time - around 200 seconds, while the new timeout ended up as liner 2 minutes. so, some users on slow systems might hit it on join. e.g. new member joins but it takes more than 2 min to sync the member and the promotion fails.
i wouldn't say it's a regression, but more of a timeout decrease.

i will send PR for 1.33 and backports to avoid the potential issue on older releases.

[neolit123]

PRs for 1.30-1.33:

• kubeadm: use the v1beta4 EtcdAPICall timeout for etcd calls kubernetes#129859
• Automated cherry pick of #129859: kubeadm: use the v1beta4 EtcdAPICall timeout for etcd calls kubernetes#129860
• Automated cherry pick of #129859: kubeadm: use the v1beta4 EtcdAPICall timeout for etcd calls kubernetes#129861
• Automated cherry pick of #129859: kubeadm: use the v1beta4 EtcdAPICall timeout for etcd calls kubernetes#129862

for 1.30, the api was technically not enabled yet, but if users complain we may have to increase the timeouts back to the 200 sec hardcoded value or more.
for 1.31+ they can use the configurable value.

[neolit123]

one user reported this here
https://kubernetes.slack.com/archives/C2P1JHS2E/p1738151866727499
probably slow infra.