fix: migrate from k8s.gcr.io to registry.k8s.io

[CIPHERTron]

This fix is a part of an umbrella issue kubernetes/k8s.io#4780

[fidencio]

@CIPHERTron, please, squash all the commits into a single one and force-push to the same branch.

[CIPHERTron]

@CIPHERTron, please, squash all the commits into a single one and force-push to the same branch.

Done! @fidencio Lmk if there are any more changes.

[bpradipt]

/lgtm
Thanks @CIPHERTron

[fidencio]

lgtm, thanks @CIPHERTron!

[fidencio]

/test
/test-tdx

[wainersm]

Hi @CIPHERTron !

This PR breaks our CI because it seems that registry.k8s.io does not contain registry.k8s.io/coredns:v1.8.6. We install k8s with kubeadm, what else should be changed to pull the right coredns image?

[fidencio]

/test

[fidencio]

Thanks for checking this out, @wainersm!

[katexochen]

Hey, I've digged a bit into this issue. The problem @wainersm mentioned isn't caused by the change of the container registry. Actually, k8s.gcr.io and registry.k8s.io contain the same images:

❯ crane ls registry.k8s.io/coredns # or `crane ls k8s.gcr.io/coredns`
...
1.2.2
1.2.3
1.2.4
1.2.6
1.3.0
1.3.1
1.5.0
1.6.2
1.6.5
1.6.6
1.6.7
1.7.0

But the coredns images were moved to coredns/coredns :

❯ crane ls registry.k8s.io/coredns/coredns # or `crane ls k8s.gcr.io/coredns/coredns`
sha256-017727efcfeb7d053af68e51436ce8e65edbc6ca573720afb4f79c8594036955.sig
sha256-8e352a029d304ca7431c6507b56800636c321cb52289686a581ab70aaa8a2e2a.sig
sha256-a0ead06651cf580044aeb0a0feba63591858fb2e43ade8c9dea45a6a89ae7e5e.sig
sha256-b82e294de6be763f73ae71266c8f5466e7e03c69f3a1de96efd570284d35bb18.sig
v1.10.0
v1.10.1
v1.6.6
v1.6.7
v1.6.9
v1.7.0
v1.7.1
v1.8.0
v1.8.3
v1.8.4
v1.8.5
v1.8.6
v1.9.3
v1.9.4

[wainersm]

Hey, I've digged a bit into this issue. The problem @wainersm mentioned isn't caused by the change of the container registry. Actually, k8s.gcr.io and registry.k8s.io contain the same images:

❯ crane ls registry.k8s.io/coredns # or `crane ls k8s.gcr.io/coredns`
...
1.2.2
1.2.3
1.2.4
1.2.6
1.3.0
1.3.1
1.5.0
1.6.2
1.6.5
1.6.6
1.6.7
1.7.0

But the coredns images were moved to coredns/coredns :

❯ crane ls registry.k8s.io/coredns/coredns # or `crane ls k8s.gcr.io/coredns/coredns`
sha256-017727efcfeb7d053af68e51436ce8e65edbc6ca573720afb4f79c8594036955.sig
sha256-8e352a029d304ca7431c6507b56800636c321cb52289686a581ab70aaa8a2e2a.sig
sha256-a0ead06651cf580044aeb0a0feba63591858fb2e43ade8c9dea45a6a89ae7e5e.sig
sha256-b82e294de6be763f73ae71266c8f5466e7e03c69f3a1de96efd570284d35bb18.sig
v1.10.0
v1.10.1
v1.6.6
v1.6.7
v1.6.9
v1.7.0
v1.7.1
v1.8.0
v1.8.3
v1.8.4
v1.8.5
v1.8.6
v1.9.3
v1.9.4

Would bump to k8s 1.26.0 fix that? Or perhaps we should be passing the registry and let kubeadm uses its default.

[katexochen]

Would bump to k8s 1.26.0 fix that? Or perhaps we should be passing the registry and let kubeadm uses its default.

@wainersm Sorry, I'm not sure what exactly the initial issue was you mentioned in #161 (comment) or where this is configured in the fist place. The logs of the CI runs don't exist anymore.