rke2-coredns-rke2-coredns became rke2-coredns

[albundy83]

Environmental Info:
RKE2 Version:

v1.27.11-rc1+rke2r1

Node(s) CPU architecture, OS, and Version:

Linux ip-10-0-1-20 6.5.0-18-generic #18~22.04.1-Ubuntu SMP PREEMPT_DYNAMIC Wed Feb 7 11:40:03 UTC 2 x86_64 x86_64 x86_64 GNU/Linux

Cluster Configuration:

3 control-planes, 2 workers

Describe the bug:

I'm not sure if it's a bug but when I move to release v1.27.11-rc1+rke2r1 (I used to be on v1.26.12+rke2r1), rke2-coredns-rke2-coredns Service name became rke2-coredns.
ConfigMap, Pods also became rke2-coredns instead of rke2-coredns-rke2-coredns.
rke2-coredns-rke2-coredns-autoscaler also rename to rke2-coredns-autoscaler.
Well, you got the idea :)

Maybe also it's the update of the rke2-coredns component as we move from 1.24.008 to 1.29.000

Sorry, I did not find more informations to determine what change this name.

Steps To Reproduce:

• Installed RKE2:

Expected behavior:

Nothing special, I just want to be sure that it's something expected

Actual behavior:

Additional context / logs:

[brandond]

when I move to release v1.27.11-rc1+rke2r1 (I used to be on v1.26.12+rke2r1), rke2-coredns-rke2-coredns Service name became rke2-coredns.
ConfigMap, Pods also became rke2-coredns instead of rke2-coredns-rke2-coredns.
rke2-coredns-rke2-coredns-autoscaler also rename to rke2-coredns-autoscaler.

Is there a problem, or are you just asking if this is expected? I suspect that this was a change to the upstream chart; we didn't intentionally alter that but it sounds reasonable.

Ref: rancher/rke2-charts@56596f9

[albundy83]

Well, as I told you, I don't know what to expect of this modification.
I'm not sure about the impact neither as I did not try it yet on a production cluster.
I discover it as my loki helm chart has this parameter set:

global:
  dnsService: rke2-coredns-rke2-coredns

and of course I add to change it to:

global:
  dnsService: rke2-coredns

because of the error:

 host not found in resolver "rke2-coredns-rke2-coredns.kube-system.svc.cluster.local." in /etc/nginx/nginx.conf:33

Yeah, I spend 1 hour understanding why my loki-gateway did not work anymore :)

My remark was more to be sure that this update was intentional and to notify users about it.

[brandond]

Looks like this came from coredns/helm#156

This can cause problems when upgrading, due to helm not sequencing the deletion of the old service and creation of the new service properly, leading to clusterIP conflicts (#5492). We'll be reverting this in our repackaging of the chart.

rancher/rke2-charts#414

[albundy83]

Sound good to me, thanks a lot :)