May
Info
GPU PaaS releases are initially rolled out via Rafay's Air Gapped Controller form factor. These will be periodically bundled and rolled out into Rafay's Production SaaS.
v3.1-29¶
29 May, 2025
Dashboard for GPU Cloud Administrators¶
Administrators and Operational Personnel now have access to fine grained dashboards of compute and app SKUs used by all their Orgs (tenants).
Data by Org (Tenant)¶
The CSP's Operational and support personnel can filter the dashboards by a customer Org by selecting the name of the Org from the dropdown.
Active Orgs¶
The CSP's Operational and support personnel can quickly determine the number of orgs (i.e. unique tenants) under management. Click on the hamburger icon (☰) to see the list of all Orgs.
Summary-Number of Compute Instances¶
The CSP's Operational and support personnel can determine the number of compute instances (i.e. compute SKUs) across all Orgs or a specific Org. The instances are presented by their status as well i.e.
- Active
- Inactive
- Failed
- Delete Failed
Users can click on the hamburger icon (☰) to see the list of all compute instances.
Summary-Number of Service Instances of Service SKUs¶
The CSP's Operational and support personnel can determine the number of service instances (i.e. app SKUs) across all Orgs or a specific Org. The instances are presented by their status as well i.e.
- Active
- Inactive
- Failed
- Delete Failed
Users can click on the hamburger icon (☰) to see the list of all service instances.
Trend- Number of Compute & Service Instances¶
The CSP's Operational and support personnel can determine the trend of compute and service instances (i.e. compute and app SKUs) for the last 30 days across all Orgs. They can filter the trend chart by profile (i.e. SKU)
Compute & Service Instances by Profile (SKU)¶
The CSP's Operational and support personnel are presented with a view of compute and service instances by profile (SKU).
Integrated Developer Tools for k8s Compute¶
End users (i.e. developers, data scientists etc) that use vCluster and dedicated Kubernetes cluster based compute instances now have access to a set of integrated developer tools to aid them with troubleshooting and visibility.
View & Download Kubeconfig File¶
End users can view and download the Kubeconfig file for their Kubernetes cluster so that they can access the cluster remotely and securely using popular tools such as the Kubectl CLI utility or desktop apps such as Lens.
Important
Secure, remote access to the virtual or standalone Kubernetes cluster using the downloaded Kubeconfig file is provided via Rafay's Zero Trust Kubectl (ZTKA)
Kubectl Browser Shell¶
End users can use the integrated browser shell to perform Kubectl operations on their virtual and standalone Kubernetes clusters remotely and securely.
Important
Secure, remote access to the virtual or standalone Kubernetes cluster using the downloaded Kubeconfig file is provided via Rafay's Zero Trust Kubectl (ZTKA)
k8s Resources Dashboard¶
End users can use the integrated developer tools to visualize and interact with Kubernetes resources on their virtual and standalone Kubernetes clusters. They can click on the They can visualize metrics for CPU, Memory, Storage etc via
Metrics Dashboard¶
End users can use the integrated developer tools to visualize metrics for CPU, Memory, Storage etc via the integrated metrics dashboards.
Cost Estimation Enhancements¶
Users are now shown the hourly rate as part of the total monthly cost estimates before they launch their compute/service instances.
3.1-28¶
16 May, 2025
Billing Cycle Enhancements¶
The billing cycle for running instances in a tenant/org will continue even if the backing profile for the instance is unshared or deleted from a project.