I enabled ACE on downward cluster from my rancher,...
# general
q
I enabled ACE on downward cluster from my rancher, but when I try to connect to the cluster, I get error indicating that I need to be logged in. When I looked at cattle-impersonation-system namespace in downstream cluster, I see only the project owner service account is configured. Is there a way configure other users configured in rancher cluster to access the downstream cluster ?
c
you need to get a new kubeconfig that contains a context for that cluster
q
Thanks. I copied the kubeconfig for the downstream cluster using rancher UI and I get the same error. Is there a way to sync the users from rancher to downstream cluster ?
c
1. what kind of downstream cluster is this, and how did you configure ACE? 2. you are sure that you’re using the context that points at the downstream endpoints?
Did you manually configure an endpoint, or did you let rancher just generate endpoints from the server node addresses?
q
The downstream cluster is RKE2. I enabled ACE by editing the yaml file. I see the context set to my downstream cluster
I think I also manually configured the ACE because enabling the ACE after editing the yaml file for cluster didn’t work. I followed steps mentioned in rancher documentation
c
right so did you set a fqdn manually, or did you let it generate endpoints based on the control-plane node addresses
are you using the -fqdn or the -node_name context to connect directly
q
I didn’t set fqdn manually.
It it using the node ip address
c
so you are using the
<CLUSTER_NAME>-<NODE_NAME>
context?
q
Yes
c
and you’ve confirmed that the webhook config file is in place on all server nodes, the configuration has been updated, and the server nodes have been restarted?
you need to do that on all the server nodes
q
I didn’t restart all server nodes. I will try that. Thanks for the help
c
that is covered in the docs
step 3
q
I need to restart only control plane nodes right ?
c
you should do it on all server nodes
q
ok. Thanks. Appreciate your help
c
you need to place the webhook config file, add the apiserver arg to config.yaml, and restart the service. on all server nodes. THEN update the cluster in the rancher UI
If you follow the documented steps it should work