SUSE Virtualization v1.8.1 Release Notes

In SUSE Virtualization v1.8, you can migrate all or selected virtual machine disks between different storage providers (such as SUSE Storage or external CSI volumes) without downtime. The guest operating system remains unaware of the storage transfer, ensuring minimal disruption to active workloads. This capability includes full UI and API support with real-time progress tracking, as well as safety checks that verify destination capacity and temporarily block conflicting operations (such as live migration) until the transfer completes.

Demo | Documentation | GitHub issue

SUSE Virtualization v1.8 introduces dynamic CD-ROM media management, allowing you to insert or eject ISO images while a virtual machine is running. Although adding or removing the CD-ROM drive device itself still requires a virtual machine reboot, mounting or unmounting ISO media within an existing drive is completely seamless. Additionally, virtual machines with mounted CD-ROM ISOs remain fully live-migratable across nodes.

Demo | Documentation | GitHub issue

Harvester v1.8 introduces a configuration setting that allows you to manually adjust Instance Manager resource reservations. If your cluster primarily uses third-party storage, you can reduce the default CPU and memory allocated to SUSE Storage. This optimization frees up cluster resources for other workloads when SUSE Storage is only needed for core system functions, such as image management and upgrades.

Demo | Documentation | GitHub issue

SUSE Virtualization v1.8 supports Pod Security Admission (PSA) to enforce Pod Security Standards (PSS). Built-in Kubernetes admission controllers enforce these standards across SUSE Virtualization system namespaces, ensuring that core workloads remain unaffected and that a seamless upgrade path exists from non-PSA to PSA-enabled images. User namespaces remain unenforced by default, preserving flexibility for you to define custom security levels or independently deploy policy engines (such as SUSE Security Admission Controller, OPA Gatekeeper, and Kyverno).

Demo | Documentation | GitHub issue

SUSE Virtualization v1.8 supports backup and restore operations for guest cluster workloads using the Harvester CSI Driver. This allows you to protect workload data by creating backups of PersistentVolumeClaims (PVCs) within the guest cluster, which are then stored using the underlying storage provider (currently limited to SUSE Storage) on the SUSE Virtualization cluster. You can restore these backups to recover from data loss or to migrate workloads across clusters, improving disaster recovery for guest applications.

Demo | Documentation | GitHub issue

The SUSE Virtualization v1.8 UI proactively alerts you when a virtual machine requires a restart to apply configuration changes, such as interface binding or resource allocation updates. If you choose to save changes without restarting immediately, a warning appears on the Virtual Machines screen to highlight virtual machines with pending changes. This prevents unapplied configurations from silently blocking future updates and ensures you can apply them at a convenient time using the UI restart action.

Demo | Documentation | GitHub issue

SUSE Virtualization v1.8 allows you to set a baseline CPU model for a virtual machine to ensure reliable live migration across hosts with different CPU generations. You can explicitly select a CPU model (default, host-model, or a named model such as IvyBridge) during virtual machine creation using the UI or Terraform Provider. SUSE Virtualization then schedules and migrates the virtual machine only to nodes that support the required CPU model, ensuring backward compatibility and preventing migration failures caused by missing CPU features.

Demo | Documentation | GitHub issue

Harvester v1.8 adds a dynamic visual topology diagram to the Virtual Private Cloud details screen, providing a clear representation of logical switches and network relationships. The diagram displays the complete network hierarchy (covering VPCs, subnets, overlay networks, and virtual machines), while reflecting real-time power states by graying out stopped or paused instances. You can navigate complex network structures using pan and zoom controls, legend toggles for node filtering, click-to-highlight node relationships, and clickable VPC peering links that redirect to associated topology views.

Demo | Documentation | GitHub issue

v1.8 introduces SUSE Virtualization-specific role-based access control (RBAC) roles within Rancher, fully enforced at the backend API level. This allows administrators to assign granular permissions tailored to specific user responsibilities. Key additions include the following:

Role permissions are additive. To maintain cluster security, sensitive cluster-level configurations (such as logging facilities) remain strictly editable by cluster administrators, with other users restricted to read-only access.

Demo | Documentation | GitHub issue

SUSE Virtualization v1.8 introduces Layer 3 connectivity support for software-defined cluster networks to serve as a dedicated underlay for Kube-OVN. Previously, inter-node virtual machine traffic traversed the management interface (mgmt-bo), causing bandwidth contention and preventing proper traffic segregation. With the latest update, you can assign a specific IP range to a custom cluster network, which SUSE Virtualization uses to automatically allocate node IPs and provision dedicated OVS bridges. This enables complete isolation of virtual machine data traffic from management traffic, improving overall network performance, security, and architectural flexibility.

Demo | Documentation | GitHub issue

SUSE Virtualization v1.8 officially supports integration with the SUSE Observability Stack Pack. SUSE Virtualization clusters can now seamlessly register with SUSE Observability to track core metrics, ensuring reliable data continuity and accurate resource reporting (such as CPU, memory, and node size) across cluster upgrades and dynamic scaling operations.

Demo | Documentation | GitHub issue

SUSE Virtualization v1.8 introduces an updated block device naming convention that removes the strict requirement for World Wide Names (WWNs). You can now attach and provision disks without WWNs across Longhorn V1, Longhorn V2, and LVM storage backends. This enhancement streamlines storage setup, particularly for NVMe drives and virtualized hardware where WWNs are often unavailable.

Demo | Documentation | GitHub issue