VMware recently announced vSphere 7.0 update 3 and this brings several new fetures in Supervisor cluster and Tanzu Kubernetes Clusters. I will help you understand some of the key features that you may be interested in too.
Key Features Addition in Supervisor Cluster
Supervisor Clusters Support Kubernetes 1.21
This release adds the support of Kubernetes 1.21 and drops the support for Kubernetes 1.18. The supported versions of Kubernetes in this release are 1.21, 1.20, and 1.19. Supervisor Clusters running on Kubernetes version 1.18 will be auto-upgraded to version 1.19, to ensure all your Supervisor Clusters are running on the supported versions of Kubernetes.
Labels and Annotations to Supervisor Namespaces
Namespaces created by DevOps users through the namespace self-service template can now have Kubernetes labels and annotations.
For more info
Edit Supervisor Cluster Configuration after Enablement
After enabling Supervisor Clusters with the vSphere networking stack, vSphere administrators can now edit the following settings from both the API and vSphere Client: Load Balancer username and password, Management Network DNS Search Domains, and Workload Network DNS servers, NTP servers, expand the service IP range, and add a new workload network. For clusters using either vSphere or NSX-T networking, you can scale up the control plane size after enablement. Note that you can only increase the scale of the cluster, reducing the scale is not supported at this time For more Information
Key Features Addition in Tanzu Kubernetes Grid Service for vSphere (TKGS)
RWX Support for vSAN Persistent Volumes
Workloads running on Tanzu Kubernetes clusters can now mount vSAN based Persistent Volumes with RWX. Refer this link for more detail
Metrics Server
Metrics Server is now included by default in Tanzu Kubernetes clusters moving forward beginning with 1.20.9+ and 1.21 Tanzu Kubernetes releases.
Ability to support No-NAT (routed) topology
Tanzu Kubernetes clusters can now be created with a networking topology that allows cluster nodes to be routed outside of the cluster network. I found this blog from Vino very helpful
Enable containerized workloads to leverage GPU acceleration on your Tanzu Kubernetes cluster
vSphere admins can now provision GPU accelerated VMs to Tanzu Kubernetes clusters, and developers can now add GPU accelerated VMs to their Tanzu Kubernetes Grid clusters with native Kubernetes commands.
Tanzu Kubernetes release (TKr) based on Ubuntu 20.04
This is our first TKr release based on Ubuntu 20.04. This image was optimized and tested specifically for GPU (AI / ML) workloads.
Modern Apps Platform Learning 