Feature Description

The CNDP allows you to configure an internal and external VIP as part of the cluster deployment. K8s uses the internal VIP while the ingress uses the external VIP to allow access to management interfaces such as Grafana, Ops-center, and Prometheus.

By default, both VIPs are part of one VRRP and failovers together. The VRRP verifies only the internal network for connectivity issues to prioritize the stability of the Kubernetes cluster. This can cause loss of connectivity to management interfaces if there’s a network failure on the external network only.

This feature adds support for splitting the VIPs into different VRRP instances to allow them to failover independently.