Hi All
Currently, I have tutor installed on an AWS Managed Kubernates cluster with version 14.x.x.
I have installed the AWS autoscaler and custom resource requests and limits modified in the deployment file to overcome the scaling issue.
The majority of the time, scalling works well, but caddy has affinity with the node for SSL certificates, which sometime lead my application into a broken/down state, because other pods already scheduled on the same node that caddy is trying to schedule and no further available memory/CPU for the caddy pod.
how can we fix this issue? Any good suggestions would be greatly appreciated.
As a work-around, we added the certificate via configmap and removed the caddy affinity, but the issue is we have to renew the let’s encrypt certificate manually.
Thanks
Kiran