Release Notes for KubeSlice EE 0.6.0

Release Date: 3rd February 2023

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 centers.

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.

Assigning Node Labels to Slices and Namespaces

To make the applications that run on worker clusters secure and keep them from being shared with other applications or clusters, node labels can be assigned to namespaces and slices. Applying node labels at a slice level applies them to the namespaces of the worker clusters that are part of the slice. To know more, see:

• Assigning node labels using YAML
• Assigning node labels using the KubeSlice Manager

RBAC Enhancements

In the RBAC YAML file configuration, a new property called cluster has been introduced to assign roles to namespaces of a given worker cluster.

Slice Gateway

With the current release, Slice Gateway spins up two pairs of gateway pods. With multiple replicas of the VPN pods, multiple gateway nodes, and multiple network connections to the remote cluster, we now have mechanisms in place to handle failure scenarios effectively.

Issues Fixed

These release notes describe the issues fixed in this version.

KubeSlice Manager

Fixed an error, which stated that the namespace is not found, was shown on the Manage RBAC page of the KubeSlice Manager when a namespace was deleted.

Known Issues

Node Labels

• Node labels erroneously entered and applied with the YAML configuration shows up on the KubeSlice Manager but these labels may not exist on the worker cluster. As a result, any pods assigned with these labels go into the Pending state.

  Ensure that the nodes are correctly labeled on the worker clusters. If nodes matching the labels configured under node affinity rules are not found, the Kubernetes scheduler places the application pods in the Pending state.

• This issue occurs when you try to remove node labels manually from a slice for a slice with existing node labels on the KubeSlice Manager's Assign Nodes page. On the Step 1 - Assign to Slices tab, removing all the node labels manually under SELECTED LABELS disables the Next button. You must at least retain one node label or use the CLEAR ALL button to remove all the node labels.

• Scaling for replicas does not pick up the new node affinity rules.

Slice Gateway Redundancy

• The Slice Gateway Redundancy is incompatible with the existing slices on the cluster. So, you must create a new slice to use the Slice Gateway Redundancy.
• Sometimes, gateway pods get stuck in the Init state due to this upstream NSM issue.
• If a node that contains the application pods and NSM webhook pod scheduled on it is restarted, the NSM containers are not injected into the application pod. This occurs as the NSM webhook pod manages the lifecycle of the NSM mutatingwebhookconfiguration.