Release Notes for KubeSlice OSS 0.2.0

Release Date 16th June 2022

KubeSlice is a cloud-independent platform that combines network, application, Kubernetes, and deployment services in a framework to accelerate application deployment in a multi-cluster and multi-tenant environment. KubeSlice achieves this by creating logical application slice boundaries which allow pods and services to communicate seamlessly across clusters, clouds, edges, and data centres.

We continue to add new features and enhancements to KubeSlice.

What’s New

These release notes describe the new changes and enhancements in this version.

Onboarding Namespaces

Namespaces that are created to run application deployments can be onboarded on a slice to form a micro network segment. Once a namespace is bound to a slice, all the pods that get scheduled in the namespace would get connected to the slice. The configuration is part of the slice YAML file. This feature onboards namespaces and not individual applications.

Breaking Change for Onboarding Applications

With the onboarding namespaces feature, onboarding each application is no longer supported. The onboarding namespaces feature can onboard complete namespaces and not individual applications.

caution

If you upgrade the worker operator to this 0.2.0 version, then the existing onboarded applications do not work as expected.

To avoid this breaking change, you must add the corresponding namespace of the existing onboarded application in the slice configuration file. To know more, see namespace isolation profile parameters.

To onboard namespaces:

1. Edit the slice configuration YAML file to add namespaces as part of applicationNamespaces . You can add namespaces in the following ways in the slice configuration YAML file:

   • Add namespaces for each worker cluster.
   • Add a wildcard * (asterisk) to add all namespaces in the worker clusters.

   info

   Ensure that the namespace that you want to onboard exists on the worker cluster.

2. Add the namespace and the corresponding clusters under the applicationNamespaces in the slice configuration file as illustrated below.

     namespaceIsolationProfile:
      applicationNamespaces:
       - namespace: iperf
         clusters:
         - 'worker-cluster-1'
       - namespace: bookinfo
         clusters:
         - '*'

   info

   Adding the asterisk (*) enables the namespace sameness, which means that the namespace is onboarded on all the worker clusters of that slice.

3. Apply the slice configuration to complete the process of onboarding namespaces.

   kubectl apply -f <slice configuration>.yaml -n <project namespace>

Namespace Isolation

By default, all namespaces on a slice are not isolated and accept traffic from any source. To secure the slice, you can selectively allow traffic to namespaces by isolating them.

The namespace isolation feature enables you to confine application namespaces with a slice. The application namespaces are isolated from other namespaces in a cluster and are connected to the slice network. This leads to the formation of a secure inter-cluster network segment of pods that are isolated from the rest of the pods in the clusters.

Intra-cluster Slice

A slice can now also be created within a single worker cluster.

Supported Kubernetes Services

This version of KubeSlice has been tested on Azure Kubernetes Service and Google Kubernetes Engine, and KIND Kubernetes clusters.

The supported Kubernetes versions for cloud clusters are 1.20, 1.21, and 1.22.

The supported Kubernetes versions for KIND clusters are 1.21 and 1.22.