-
Notifications
You must be signed in to change notification settings - Fork 2.9k
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
[BUG] [Instability] ACI-CNI variables not configurable using Rancher server UI #44980
Comments
@rancher/rancher-team-2-hostbusters-qa Need help with reproducing this issue. I am trying to reproduce with |
Able to reproduce the issue. Steps to reproduce on Rancher v2.8.3:
Notes:
|
Debugging notes:
Rancher v2.6 / RKE release/v1.3
Rancher v2.7 / RKE release/v1.4
|
We've confirmed the root cause of this issue. rancher/webhook is vendoring an older version of RKE so the fields disappear if webhook doesn't have the right version of RKE. This also explains why editing cluster CRs via kubectl doesn't work because these requests also go through webhook. Rancher v2.8.3 vendors RKE v1.5.7 and runs webhook v0.4.3 which vendors RKE v1.5.7-rc2. The new ACI field |
Validated that the ACI values do not disappear after provisioning the rke1 cluster on 2.8.4-rc4 |
Rancher Server Setup
Information about the Cluster
User Information
What is the role of the user logged in?
Describe the bug
We are experiencing difficulties where the values we input for ACI fields are being discarded, and the default values are being used instead. This issue seems to be with the Rancher server. When we update the cluster configuration in the UI by inputting fields in the "Edit as YAML" option, the values we provide are being discarded/disappeared upon clicking the save button.
On the API level, the post request is sent out with our fields intact, but the response does not include them.
Also, the variables we added for 6.0.3.1 (RKE 1.4.9 which were successfully tested on Rancher 2.7.9), are not being picked up.
The variables introduced with ACI-CNI 6.0.4.1 are not being picked in by Rancher server v2.8.3.
To Reproduce
1.Go rancher server UI , Create a Custom rke1 cluster
2. Give some name and select k8s version you want (v1.26.11-rancher2-2)
3. Go to the top, Select ‘Edit Yaml’ option
4. Copy the generated network provider file contents in the network section.
Add the custom values for aci cni as below:
Edit: for v2.8.3 add the following fields:
Result
We can verify aci values after creating cluster are missing by below:
a. Go to 'Edit Config'
b. Edit As YAML
c. Recheck the above added aci variables are present there.
we see missing aci values now:
Also if everything works fine then, we should be able to get these values from kubectl commands like:
kubectl describe deploy -n aci-containers-system aci-containers-controller | grep toleration
Edit: for v2.8.3, just the 6.0.4.1 variables are being discarded:
Expected Result
Once the aci cni fields added in in Network section using 'Edit As YAML', it should reflect on Kubernetes resources and in UI's 'Edit As YAML' option as well.
(edited to include details pertaining similar issue on Rancher v2.8.3. separate issue created here: #45200)
The text was updated successfully, but these errors were encountered: