This message was deleted.
# rke2
a
This message was deleted.
c
Those docs are for Rancher, you should be looking at the RKE2 docs. https://docs.rke2.io/install/containerd_registry_configuration#configs
s
Thanks @creamy-pencil-82913! Do you know if I can apply this to already running cluster? Do I just restart the rke2-server?
I am running a downstream cluster with 3 etcd nodes and 2 ctrl nodes and 3 worker nodes. It recommends doing this on each node, but which type of node?
c
yes you can change that on a running cluster. You need to configure the registries.yaml on all of the nodes in the cluster.
This is just passed through directly to containerd running on that node, so you need to configure it on any node you want to trust that registry. So probably all of them?
If you provisioned the cluster via Rancher there’s actually a spot in the Cluster management UI for configuring registries, and it will push it out to all the nodes for you.
s
I just found that I can configure this directly from rancher UI for the cluster config. It seems to write the registries.yaml file to all nodes.
c
yep thats what I was just referring to
s
it still issues, but maybe I am not using the registries.yml correctly.
Copy code
{
   "configs":{
      "harbor10165":{
         "auth":{
            "username":"my-username",
            "password":"my-assword",
            "auth":"",
            "identity_token":""
         },
         "tls":{
            "ca_file":"",
            "cert_file":"",
            "key_file":"",
            "insecure_skip_verify":true
         }
      }
   },
   "mirrors":{
      "harbor10165":{
         "endpoint":[
            "<https://harbor10165.senode.dev>"
         ]
      }
   }
}

kubectl run webapp --image=harbor10165.senode.dev/sgs/webapp:2.0 --port=8080 --expose=true --labels="app=webapp"
The pod fails with x509
corrected the hostname and now the x509 is resolved.
Thanks again!