https://rancher.com/ logo
Join Slack
Powered by
This message was deleted.
# rke2
a
This message was deleted.
c
did you follow the upstream docs to tag everything? nodes, subnets, and so on?
a
🤔 node are automatically tagged right ?
Copy code
<http://kubernetes.io/cluster/c-m-bt79wfmm|kubernetes.io/cluster/c-m-bt79wfmm>	owned
Name	test-pool1-68e23cd7-2hq8k
but subnets not ect. You are right.. i was following: https://aws.amazon.com/blogs/containers/exposing-kubernetes-applications-part-1-service-and-ingress-resources/
Is there any impact on tagged resources if i delete the cluster ? My subnets are shared with our infrastructure.
c
by default nothing is tagged in EC2
you must have some other automation that is handling that for the instances
or maybe rancher does it, IDK. RKE2 doesn’t.
a
Just checked my instances
👍 1
Just found the doc related: https://ranchermanager.docs.rancher.com/pages-for-subheaders/use-existing-nodes#3-amazon-only-tag-resources Thank you. How 
shared
and 
owned
are used by Rancher ?
c
they are not. they are used by AWS cloud and load-balancer controllers.
a
I don’t want my subnet to be deleted if the cluster is deleted. I suppose shared is to prevent that no ?
c
I don’t believe Rancher will delete the subnets no matter what you do
these are all just for the AWS controllers.
🙏 1
a
After losing one afternoon to understand, i hope it was the reason. There is no way to found the reason in logs ?
c
I mean, you could check? Did you check the alb controller logs?
a
I tried but i was not able to found something. I’m novice on Rancher and Kubernetes.
ALB is not even created on AWS.
c
no, the controller pod logs
you said you deployed it, right?
a
i deployed the nginx-ingress
c
ah. I’m sorry, you said you wanted to expose it with an NLB so I assumed you were using the AWS LB controller
Did you instead deploy another ingress-nginx controller? RKE2 already comes with ingress-nginx.
a
I deployed the Cluster without the option and i deployed it on my own
c
why? that’s going to be more work, especially if you don’t know much about Kubernetes
you’d be better off just using the default ingress-nginx and pointing your NLB at ports 80 and 443 on the nodes, if you want to do it by hand,.
a
I don’t want to maintain on my own Targets Groups ect.
c
then you need to deploy one of the AWS load-balancer controllers, and have that do it for you.
and then configure the ingress controller to use a load-balancer service instead of a daemonset
a
Exposing NGINX Ingress Controller via a Load Balancer
there is a lot of possibility 🥲
nginx-ingress seems to be a more generic solution if i move from AWS
c
ingress and load-balancer controllers are different things
✅ 1
you generally want a load-balancer in front of your ingress
✅ 1
load-balancer is layer 3 (tcp), ingress is layer 7 (http/https)
a
I think i don’t undersand the difference between: AWS Load Balancer Controller and Nginx-Ingress-Controller.
c
sounds like you have some studying to do!
a
yes 😅
c
a
Back on the issue...
I tagged my security groups, subnet and instances. And the service is still stuck on Service is ready:Load balancer is being provisioned
Fixed 🎉
18 Views
Open in Slack
PreviousNext
Powered by