When installing KUMA in a fault-tolerant configuration, the file named artifacts/k0s-kubeconfig.yml is created in the installer directory. This file contains the details required for connecting to the created Kubernetes cluster. The same file is created on the main controller in the home directory of the user set as ansible_user in the inventory file.
To ensure that the Kubernetes cluster can be monitored and managed, the k0s-kubeconfig.yml file must be saved in a location available for the cluster administrators. Access to the file must be restricted.
Managing a Kubernetes cluster
To monitor and manage a cluster, you can use the k0s application that is installed on all cluster nodes during KUMA deployment. For example, you can use the following command to view the load on worker nodes:
k0s kubectl top nodes
Access to the KUMA Core
The KUMA Core can be accessed at the URL https://<worker node FQDN>:<worker node port>. Available ports: 7209, 7210, 7220, 7222, 7223. Port 7220 is used by default to connect to the KUMA Core web interface. Access can be obtained through any worker node whose extra_args parameter contains the value kaspersky.com/kuma-ingress=true.
It is not possible to log in to the KUMA web interface on multiple worker nodes simultaneously using the same account credentials. Only the most recently established connection remains active.
If you are using an external load balancer in a fault-tolerant Kubernetes cluster configuration, the ports of the KUMA Core are accessed via the FQDN of the load balancer.
Page top