-
Notifications
You must be signed in to change notification settings - Fork 4.7k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Kops cluster-autoscaler addon not getting deploy up #11759
Comments
Are you able to ssh into a control plane node and run |
I've checked the The only other thing that I found could be related is these lines about the channels, which happened once in a minute:
I've also tried to do the same operation on a brand new cluster created from scratch for the test, and it worked well. I guess the problem is with something in this specific cluster, which might not be correctly updated from previous kops/k8s versions. When I run |
None of the lines above match the CAS addon (v1.19.{0,1}, id=k8s-1.15), so it looks like it wasn't added. maybe a typo in your manifest? |
Do you mean the cluster manifest? The cluster manifest was accepted by kops without any complaint. I was expecting it to complain if the cluster manifest is wrong. Also I've checked it, and it looks fine. the clusterAutoscaler section is almost a copy and paste from the addon docs. |
Oh! Wait, last change I did in the cluster was to remove the section. adding it back now. |
Done, now I got these two new lines:
|
As of #10246, I've checked my kube-system annotations, but I barely have an idea of what this should look like:
|
Should I try to remove the cluster-autoscaler addon folder from my bucket for this cluster? |
Yeah I noticed a bug here. |
The workaround worked well, thank you! |
Thanks @olemarkus |
Is there a way that we can add or direct cluster autoscaler to use aws-use-static-instance-list= true in cluster spec? If I am editing the deployment manually and adding command line parameter then its working but I want cluster spec to do it |
Got the solution, we can add specific image so in 1.16.4 static instance list is supported by default |
/kind bug
1. What
kops
version are you running? The commandkops version
, will displaythis information.
I've tried this with both kops 1.19 and kops 1.20.
2. What Kubernetes version are you running?
kubectl version
will print theversion if a cluster is running or provide the Kubernetes version specified as
a
kops
flag.I'm using kubernetes 1.19.9.
3. What cloud provider are you using?
AWS, on EC2.
4. What commands did you run? What is the simplest way to reproduce this issue?
On a cluster provisioned on EC2 using kops with manually created
deployment for cluster-autoscaler, I removed the deploy. Then edit the
cluster definition to include the cluster autoscaler example
configuration from the documentation. I run
kops update cluster example.k8s.local --yes
and thenkops rolling-update cluster example.k8s.local --yes
.5. What happened after the commands executed?
The cluster is rolling updated, but there's no cluster autoscaler pods
not deployment.
6. What did you expect to happen?
I expected to have the cluster-autoscaler deploy created and
corresponding pod up and running on the kube-system namespace.
7. Please provide your cluster manifest. Execute
kops get --name my.example.com -o yaml
to display your cluster manifest.You may want to remove your cluster name and other sensitive information.
8. Please run the commands with most verbose logging by adding the
-v 10
flag.Paste the logs into this report, or in a gist and provide the gist link here.
Update: https://controlc.com/ca90e288
Rolling-update: https://controlc.com/58c59773
9. Anything else do we need to know?
I've tried forcing a rolling update, but the result is the same, no
deploy for
cluster-autoscaler
. The verbose output for the forcedrolling update is too big to push, but I can send them. I've checked
kubelet's logs but found nothing about the cluster autoscaler
manifest.
The text was updated successfully, but these errors were encountered: