Open cezhang opened 1 year ago
it seems the bootstrap kubeconfig has problems, would you try to expose the bootstrap kubeconfig from the managed cluster bootstrap kubeconfig secret kubectl -n open-cluster-management-agent get secrets bootstrap-hub-kubeconfig -oyaml
and use it to connect the hub
or would you try the command clusteradm https://github.com/open-cluster-management-io/clusteradm to registration your clusters, I think the clusteradm is more easy than make scripts, thanks
or would you try the command clusteradm https://github.com/open-cluster-management-io/clusteradm to registration your clusters, I think the clusteradm is more easy than make scripts, thanks
I was using clusteradm. Already try twice, still the same problem
which version of clusteradm are you using? we are testing k3s and update the doc here https://github.com/open-cluster-management-io/open-cluster-management-io.github.io/pull/318/files#diff-d98b6f00dd920af6592871b061781aaa7e9e350aa1eafaf6ee62f689bf43a4e2
which version of clusteradm are you using? we are testing k3s and update the doc here https://github.com/open-cluster-management-io/open-cluster-management-io.github.io/pull/318/files#diff-d98b6f00dd920af6592871b061781aaa7e9e350aa1eafaf6ee62f689bf43a4e2
client version :v0.4.1
server release version :v1.24.4+k3s1
default bundle version :0.9.1
hrm I think the failed case is running AKS as a hub (or ACK?). We met some issue by running aks as hub before. Will recheck and come back later.
yes, Azure k8s service as hub
OK, we will try it on our side, thanks
/cc @ldpliu
@cezhang I tried aks(with default config) in my env, and it works fine. Do you have any special config like CNI when you create the aks cluster?
I try it in a new aks env(1.25.1), it's ok, but not with my 1.22(aks), all with default config. Let me take a deep look in this before I get back here.
BTW, thanks.
Hi, I follow the
Quick start
section, and do that in two scenario :logs show in
klusterlet-registration-agent
inopen-cluster-management-agent
namespace:I also check the log from
cluster-manager
Same in both scenario
What am I missing?