Selfmanaged Kubernetes

How does it benefit me?

Getting a Selfmanaged cluster means you get a vanilla K8s cluster where you are in complete control over your own cluster. You manage everything right down to the core of your cluster. This gives you real freedom where you can choose where to run your core components such as your etcd cluster.

This configuration needs experience and knowledge on how to manage, run and operate a k8s cluster.

Baremetal nodes

Each node is configured to your needs, and can come in any hardware configuration. Nodes can have more and/or bigger disks or a JBOD connected to create an internal Ceph cluster inside K8s. Every cluster is built with a minimum of 6 nodes -- 3 masters and 3 workers.

Nodes are built in cooperation with you and each hardware component is carefully chosen and tested to match your needs.

ASERGO services

When you get a k8s cluster you will still get all the known benefits known from our dedicated series, like access to DNS servers, Availability Zones, storage and access to our network spanning over 3 data centers (2 in Denmark, 1 in Germany).

A cluster is locked to only operate in 1 data center, but it is possible to split up the cluster and make it operate across 2 or even 3 data centers.

Most of these functions can be accessed through the ASERGO Dashboard.

Support

ASERGO has a dedicated k8s team that works with production clusters on a daily basis and stand by to help you with anything related to your k8s cluster.

Quality of life Addons

A cluster from ASERGO is a vanilla cluster included with a few addons to help you get started using the cluster in no time. Addons include a central pod logging and a web UI to manage the cluster.

The web UI can be used to deploy new apps and troubleshoot any cluster problems. It also provides an opportunity to modify already defined deployments, jobs, daemonSets and StatefulSets ect.

The web UI also provides live information stats of the cluster and its nodes, such as RAM, CPU and volume usage.

The login installed on the cluster is an EFK stack (Elasticsearch, fluentd and Kibana) to centralize all pod logs. Custom filters can be setup in fluentd's ConfigMap to match your apps log output format.

This ConfigMap can either be edited from the web UI or from cli with tools such as kubectl.