Table of Contents
Well, well, well, KubeCon + CloudNativeCon EU 2025 has come to an end, and we made so many friends and memories!
Let's review some of the exciting things our team did at KubeCon:
Platinum Sponsors at Platform Engineering Day: The buzz around platform engineering and internal developer platforms remains strong. We saw increasing interest from attendees looking to better understand the landscape and searching for tooling to build their ideal platform. At the vCluster booth, we were thrilled to share insights on creating your internal Kubernetes platform with vCluster at the core. We encourage you to join our Slack community to share your platform stories and discuss how you're building your own internal Kubernetes platform.
Gold Sponsors at KubeCon EU: As Gold sponsors, our booth was bustling with great conversations! Key topics discussed included:
- How vCluster helps with multi-tenancy.
- Using vCluster for internal developer platforms.
- vCluster for ephemeral environments.
- Deep dives into specific vCluster features.
One of the biggest takes from our conversations were
- Bare metal Kubernetes - There was also a very intense interest in our new product vNode, particularly in the ability to fully support bare metal Kubernetes. Many enterprises worldwide running Kubernetes on VM's are looking to replace their traditional server virtualization technology with a new virtualization layer (e.g., KubeVirt). However, with our announcement at Kubecon of our new product vNode, enterprises can instead greatly simplify their architecture by running Kubernetes fully on bare metal and then creating multiple virtual Kubernetes clusters, saving money, reducing risk, and accelerating TTM (time to market).
- GPU sharing - Another common discussion at Kubecon was the ability to use vCluster for GPU sharing. As discussed by CoreWeave in their article here, vCluster and bare metal Kubernetes allows a better architecture to share GPUs. Now with vNode and vCluster, several major enterprises were very interested in initiating discussions on vNode and vCluster.
The overall vibe indicated that vCluster has matured significantly, with attendees coming prepared with detailed use cases and questions.
In-Booth Demos: Our very own Kurt and Saiyam delivered some impressive live demos. Kurt demonstrated running a virtual cluster on bare-metal Raspberry Pis, a setup where attendees could visually see exactly where their clusters were running. Saiyam conducted multiple demos each day, showcasing vCluster for shared platform stacks, internal Kubernetes platforms, and ephemeral PR environments.
Book Signings and Giveaways: We hosted book signings with amazing authors:
- Mauricio Salatino (Senior Software Engineer II, Diagrid) – "Platform Engineering on Kubernetes"
- Artem Lajko (CloudOps Engineer, iits-consulting) – "Platform Engineering on Kubernetes"
We also gave away 50 copies of Marc Boorshtein’s book "Kubernetes – An Enterprise Guide" and held several raffle giveaways that attendees greatly enjoyed.
Talks from the Team: Team LoftLabs delivered impactful sessions at the main KubeCon event:
- Saiyam Pathak and Paco: In their session, "A Huge Cluster or Multi-Clusters? Identifying the Bottleneck" – they discussed best practices for managing large Kubernetes clusters, highlighted challenges like multi-tenancy, and demonstrated how solutions like vCluster effectively address these issues. Over 1500 attendees registered, and the talk received excellent feedback.
- Scott McAllister: In his session, "Are You Covered? Falling in Love With E2E Testing" – Scott explored end-to-end testing strategies for Kubernetes, emphasizing GitOps and ArgoCD. He also illustrated how vCluster simplifies these processes, earning positive responses from the audience.
Announcements at KubeCon
We came to KubeCon with bunch of new stuff:
vNode : We at KubeCon launched vNode to address a critical gap in the ecosystem: the trade-off between workload isolation and resource efficiency at the node level. vNode enhances Kubernetes multi-tenancy by providing lightweight, secure isolation of tenant workloads directly on shared physical nodes using Linux user namespaces, eliminating the need for costly separate nodes or complex micro-VM solutions like Kata Containers or gVisor. Designed for platform engineering teams in regulated environments, resource-intensive setups like AI/ML, or scenarios with noisy neighbor issues, vNode offers strong isolation, low overhead, full tenant autonomy for privileged workloads, and seamless Kubernetes-native integration across major clouds. Paired with vCluster, which isolates Kubernetes control planes, vNode ensures comprehensive multi-tenancy by securing the node layer, optimizing both resource use and security. We invite users to join the private beta at vNode.com to explore this innovative solution, marking their next step in advancing cloud-native infrastructure at scale.
vCluster 0.24
The vCluster v0.24 release introduces two major features: Snapshot & Restore and significant enhancements to Sleep Mode. Snapshots are now built directly into vCluster, providing quick, lightweight backups that can be easily exported and restored from OCI registries, S3, or local storage. This simplifies tasks such as reverting virtual clusters, migrating configurations, cloning clusters, or moving between Kubernetes clusters. Additionally, v0.24 unifies the previous vCluster-native and platform-based Sleep Modes into a single, streamlined solution that automatically pauses workloads and, if using an agent, the control plane after periods of inactivity. Other improvements include an enhanced Kube-Config export feature allowing multiple secrets, deprecation notices for certain legacy images, and updated namespace reuse policies.
Rancher OSS integration
We announced the Open Source vCluster Integration for Rancher, enabling users to easily create and manage virtual clusters directly from the Rancher UI, similar to traditional Kubernetes clusters. This integration, initially requested by Rancher users, was previously only available commercially but is now released as a fully open-source solution. Built as a standalone Rancher UI extension accompanied by a lightweight controller, it simplifies deployment and management significantly. Users simply install the operator and UI extension, grant permissions, and can immediately offer self-service virtual clusters within Rancher. The integration manages user permissions automatically, providing a seamless experience. While some advanced features, such as project permissions and SSO synchronization are still in development, the current solution is entirely free, open-source, and ready to use, offering the simplicity of self-service namespaces combined with the power of full virtual clusters.
What's Next? The LoftLabs team is already busy adding more features to vCluster and our platform to help users further optimize costs. Our sales team will follow up with everyone we met at KubeCon for more detailed demos. If you missed connecting with us, reach out via Slack we’d love to discuss how vCluster can address your specific needs.
We'll continue onboarding participants for the vNode beta program, prioritizing our existing customers, to explore advanced isolation using the vCluster and vNode combination.
Meanwhile, our DevRel team will continue to rock the stages at conferences and meetups globally, including Gitex Singapore, KCD NYC, and local community events. Join our Slack community and explore our engaging workshops based on real-world scenarios!
Workshops
- Ephemeral Pull Request Environments Using vCluster
- Shared Platform stack using vCluster
- How to build cloud native platforms with vCluster
Star the vCluster repo if you like the work we are doing!
