Closed Elektronenvolt closed 7 months ago
Same here. I can confirm that it is a Bug. If you have installed 2022 initially, it brakes the Update. It will look for the not existing 2019 image and the Update gets stuck.
After a couple of hours it will timeout the update. We had to create a new nodepool :(
Has this been testet before publishing the release?
It never broke the update in our case, just changed the OS image from WS 2022 to 2019 with the update.
And, I've noticed it the first time with the previous release (August), but I was not sure if a bug or somebody changed the nodepool on purpose. With the October release I've tested for this on dedicated test cluster where only I have access to.
Hello @Elektronenvolt, we are in the process of pushing a new PowerShell Version with the fix for this issue. Thank you for reporting it. We are also adding additional validation to our pipelines to ensure we will not miss similar issues in the future.
Hello @Elektronenvolt, we are in the process of pushing a new PowerShell Version with the fix for this issue. Thank you for reporting it. We are also adding additional validation to our pipelines to ensure we will not miss similar issues in the future.
Does that mean it is fixed now? Update the PS Module?
Describe the bug After Updating the Kubernetes version of cluster node pools - Windows Server 2022 nodes are changed to 2019.
To Reproduce Execute:
Update-AksHciCluster -name "cluster" -kubernetesVersion "version"
Expected behavior Don't change the OS version.
Screenshots
Environment (please complete the following information):
Additional context I've seen this issue already with the previous August release, but was not sure if somebody changed the node OS on purpose. For this release I used a separate test cluster where I tested the updates.
Collect log files