DevConf.CZ 2020

When it comes to managing a Kubernetes cluster, we will want to perform maintenance operations on the underlying nodes from time to time.
This may include package updates, kernel upgrades, hardware repair and other voluntary distributions.
In this talk we will learn about:
- How to gracefully drain a node from its existing workloads using Kubernetes API.
- What should be taken into account for heavy workloads such as VMs and storage.
- Avoiding outages using pod disruption budgets.
- The concept of server side drain.