Webinar: Kubernetes and High Availability for Services and Control Plane

by Ricardo Brito Da Rocha (CERN)

Wednesday 20 May 2020, 14:00 → 15:00 Europe/Zurich

https://cern.zoom.us/j/94143542671 (Zoom)

Recording: CERN Video YouTube

Join us in zoom (Meeting ID: 941 4354 2671). Recording will be available shortly after the session.

This session will cover the multiple aspects affecting Kubernetes high availability, both for the cluster's control plane and applications.

We will focus on areas that are generic and some that are specific to our deployment, including:

• How to enable multi master clusters, and the impact on control plane availability
• Selecting node anti-affinity to spread workloads within the same availability zone (AZ)
• Using node groups to have clusters spreading multiple AZs
• Relying on ReplicaSets and Pod affinity definitions to get highly available applications
• Setting disruptions budgets to guarantee availability during application upgrades

 

 

 

 

 

 

 

nginx-affinity.yaml

nginx-ingress.yaml

nginx-service-lb.yaml

nginx-service.yaml

nginx-topology.yaml

Webinar Kubernetes High Availability Services and Control Plane.pdf