Table of Contents
We have been working on something new, and it is finally time to share it with everyone. Our latest release is vCluster.Pro. vCluster.pro expands on vCluster by adding enterprise-ready features and support.
vCluster has been out now for a couple of years and has provided over 40 Million virtual clusters around the world. As one of the most popular open-source projects from Loft, we have steadily added features and value with each release. Enterprises wanted more, and we listened. We are excited for everyone to try it out.
#Enterprise Grade Virtual Kubernetes
Optimization and Enterprise-level features were the driving force behind creating a pro version of vCluster. The first release is all about optimization and empowering our users while also providing the support needed to reach the kind of scale required for large-scale critical deployments.
If you want to utilize your worker nodes optimally, the question of overhead will come up when asking about any new addition to your stack. How many resources will we need to dedicate to run this? With vCluster.Pro you can run an isolated control plane in another cluster while your workloads run on the hardware that was built to support those workloads.
A popular use case would be training or inference workloads. Ensuring that your GPUs and hardware are dedicated to only running training and inference and moving other less critical control plane pods to a node without a GPU can increase productivity, especially at scale.
vCluster.pro is the only platform for creating and managing virtual Kubernetes clusters at scale.
vCluster.Pro integrates the CoreDNS container and the vCluster syncer. As a result of this, when enabled, there will only be a single pod containing all 3 components.
Key benefits are:
Avoiding Cluster Max Pods Limit: As clusters scale, the number of pods allowed or kubernetes API server load may be a bottleneck.
Optimized Resource Allocation: More components sharing the same allocated resources instead of having independently allocated resources means less wastage of resources due to bin-packing constraints.
#Isolated Control Plane
The Isolated Control Plane provides a way to separate the control plane from the worker nodes where the workloads are running. Headless mode basically means – no deployment or statefulset will be created for the vCluster. This feature allows you to deploy the vCluster control plane in one cluster and the workloads into another.
The name of the game is optimizing what is running on your critical infrastructure and moving the administrative stuff to a less important cluster.
The Sync Patch feature allows you to patch Kubernetes resources during the sync process between the virtual cluster and the host cluster.
vCluster.Pro uses a Chainguard busybox image as its base image in place of Alpine to maximize security. This means that known vulnerabilities are patched very quickly. Employing non-root images introduces an additional tier of security, making them highly recommended for production deployments.
Obviously, the most important thing with any pro offering: a UI! Now, a user interface can be used to view details about your virtual clusters. But wait, there’s more!
#Upgrade from Open Source
Bring your virtual clusters with you. With vCluster.Pro you can import and upgrade the virtual clusters you are already using.
#Controller & CRD
With vCluster.pro, you can manage your virtual clusters with Custom Resource Definitions. Provisioning virtual clusters is even easier and can be done via kubectl, Terraform, CI Sytems, GitOps, vCluster CLI, and the UI.
Normally, allowing a Self-Service model with Kubernetes would probably scare most platform teams. What if you could limit the spending by building on top of your already provisioned Kubernetes clusters? With Virtual Clusters, you aren’t spinning up new hardware to support a new user, you’re creating a new deployment on top of your already existing clusters and providing a new API endpoint for them.
With vCluster.pro, you get an entire platform dedicated to providing a GUI, and CLI, to users who just want to create resources, copy the auth details, and get to work. Manage your users and teams and provide them access to projects that can further split up what each user can see. Limit the ability for users to remove resources that are owned by another team or enable collaboration by granting permissions to multiple teams in the same project.
With sleep mode, you can put virtual clusters to sleep, which means that vCluster.pro will set replicas: 0 for all replica-controlled resources such as Deployments and StatefulSets inside the virtual cluster. This means that Kubernetes will delete all pods, but the entire configuration of resources within the virtual cluster remains.
Virtual cluster templates are, as the name implies, templates that you can create, which can then be applied when creating a virtual cluster. These templates can be used to configure any virtual cluster settings that you would configure on an individual virtual cluster.
Virtual cluster templates are used for creating a standard developer environment for a given team or organization or to template out pre-built environments to be used in continuous integration and testing.
A cluster is a Kubernetes cluster that can be connected to and managed from the UI. This can be a local Kubernetes cluster where vCluster.Pro is installed or in an external cluster deployed on a cloud provider.
Once connected, you can create virtual clusters in the selected cluster.
What would a pro offering be without amazing support to go along with it? With vCluster.pro, you receive SLAs for Response and Resolution, Enterprise-Only Support Channels, and access to the makers of vCluster.