Open Source Summit Japan & Automotive Linux Summit 2021

Today, many companies are using Kubernetes in production environments, but there are some big issues in Kubernetes. One of the issues that we are facing with our customer’s project is that pods with PV fail to migrate when a Kubernetes node is down.

When a worker node in a Kubernetes cluster goes down, the controller will try to recreate pods on a different node, but currently, it will fail for recreating pods if the pods were attached with persistent volume (PV) because Kubernetes can't detach PVs attached pods on failed nodes. Many of companies do force deletion of pods; in this case, as a workaround in this situation, but originally, pods should be deleted automatically in this case.
In order to fix the issue, Github issues/Pull Requests/KEPs (Kubernetes Enhancement Proposals) have been posted many times by operators/developers. But unfortunately, most of them have been closed because discussions were not continued.

Now, we are discussing how to fix this issue on the KEP posted in 2019 and issue, but it seems to take a while to fix the improvement plan. In this session, I will introduce the issue and current discussions, and also, I will talk about the difficulty of development in the open source community and what we are doing/did as a solution.