Ravi Tripathi:
@Tej_Singh_Rana one question i have, once any pod reaches its CPU limit it is going go restart a pod or terminate a pod. Pod is configured with liveliness probe. How to debug that, like which error it will through when cpu limit cross.
Mumshad Mannambeth:
Hi Ravi, once a pod hits CPU limit it will be throttled. meaning it can’t use any more CPU. It won’t be restarted. If a pod(container) hits memory limits then it will be killed. I believe we discuss this in the resource limits section of the course.
Ravi Tripathi:
@Mumshad Mannambeth @Tej_Singh_Rana Yes I have gone through that, but I am facing a prod issue which very strange to me.
Pod are getting restarted what I feel like liveliness probe is doing that, but I tried to find error regarding CPU limit reached or throttled. I have not found that, not sure why this happen. Eventually increase CPU limits and issue get resolved.
I just want to know steps that I can use to debug this kind of issue.
Mohamed Ayman:
@Ravi Tripathi you should have a log solution for the prod. Also, you can check this useful guide https://kubernetes.io/docs/tasks/debug-application-cluster/debug-application-introspection/