Cluster information:
Kubernetes version: 1.16
Cloud being used: AWS/EKS
Host OS: Windows
CNI and version: amazon-k8s-cni:v1.7.5
When I get logs of Windows pods or exec into Windows pods I often receive timeout errors such as the below:
Error from server: Get https://10.x.x.x:10250/containerLogs/namespace/pod-id dial tcp 10.x.x.x:10250 i/o timeout
It only happens some of the time and often even when there is almost no resource usage on the Windows node. This never happens on Linux nodes. Does anyone else run into these timeouts on Windows nodes? Does anyone have ideas as to the cause?
Thanks for any help you can provide.