Proxmox Virtual Environment (VE) 9.0 Beta marks a significant evolution of the open-source virtualization platform. Built on Debian 13 “Trixie” with Linux Kernel 6.14, it introduces critical new features and backend improvements that expand its use cases across modern datacenter infrastructure.

The Proxmox VE 9.0 Beta release focuses on enhancing storage flexibility, software-defined networking, Ceph integration, and infrastructure-level resilience, while deprecating legacy components such as GlusterFS and the VM.Monitor privilege.

Base System Upgrades

Debian 13 “Trixie” Foundation

Proxmox VE 9.0 runs on Debian 13, providing a more secure, modern base with improved kernel modules, better hardware support, and updated libraries. It also means better compatibility with next-generation AMD and Intel processors, networking chipsets, and storage interfaces.

Kernel 6.14.8-1

The included kernel version brings improvements for:

• Intel/AMD performance optimizations
• ARM64 support improvements (for edge deployments)
• Up-to-date drivers for RDMA, NVMe, and PCIe devices

Virtualization Stack Upgrades

QEMU 10.0.2

The virtualization engine QEMU is updated to 10.x, bringing:

• Faster live migration performance
• Better NUMA and CPU topology awareness
• Support for new guest features (e.g., Secure Encrypted Virtualization)

LXC 6.0.4

The updated LXC container engine aligns with current standards and improves:

• Resource isolation (cgroups v2)
• Network namespace support
• Compatibility with unprivileged containers

OpenZFS 2.3.3

This introduces several major ZFS enhancements:

• RAID-Z Expansion: Dynamically add vdevs to RAID-Z pools without rebuilding
• Improved I/O pipeline and metadata handling
• Snapshot space estimation and sync performance tuning

Advanced Software-Defined Networking (SDN)

One of the most anticipated features, Proxmox VE 9.0’s SDN system now supports multi-layer fabrics, ideal for spine-leaf topologies or distributed Ceph storage networks.

New SDN Features Include:

• Fabric-aware routing across clusters
• Advanced VLAN/VRF management
• OVS (Open vSwitch) enhancements for L3 routing
• Better GUI/CLI integration for SDN topology visualization

Ceph Squid 19.2 – Default Storage Backend

With Ceph Squid 19.2, Proxmox moves forward from the older “Reef” and “Quincy” versions. Key benefits:

• Better memory efficiency and IOPS
• Improved RBD/NFS Gateway performance
• More consistent multi-tenant support
• Enhanced CephFS support for shared virtual disk scenarios

This makes Squid the preferred choice for:

• Hyper-converged infrastructure
• Multi-node Proxmox + Ceph clusters
• VM high availability with fast live migration

Storage System Improvements

LVM Thick-Provisioned Snapshot Support

LVM-backed storage (especially for Fibre Channel or iSCSI SANs) now supports snapshots, a long-missing feature in thick provisioning scenarios. This allows:

• Scheduled backups and rollback on thick volumes
• Snapshot-based VM cloning
• Live VM backup without downtime

Storage Migration Enhancements

Storage migration operations have been optimized for:

• Live VM disk movement across nodes or pools
• Improved disk format conversion (e.g., qcow2 → raw)
• Parallel I/O scheduling with minimal VM I/O disruption

Web Interface and Usability Enhancements

Proxmox VE 9.0 beta includes several GUI improvements:

• Improved network graphs with live SDN statistics
• Better error/warning handling during backup/migration
• VM/CT overviews with collapsible resource sections
• Smart wizards for SDN, Ceph, and HA setup
• Inline detection of misconfigured firewalls, broken NFS mounts, or stale nodes

Security, Permissions & Policies

Deprecated: VM.Monitor Privilege

The VM.Monitor role has been removed. Monitoring access now routes via:

• Sys.Audit for global read access
• Specific guest agent permission scope (fine-grained control)

The new permissions model aligns better with:

• Multi-tenant setups
• API-driven VM orchestration tools (Terraform, Ansible, etc.)

Improved NIC Naming Strategy

With kernel 6.14+, some NIC names change (e.g., eno1 → enp1s0). Proxmox now:

• Auto-detects renamed interfaces
• Maps them in the installer and boot scripts
• Warns users during upgrade (via pve8to9 checklist)

Removed or Deprecated

GlusterFS Support Removed

Due to upstream GlusterFS instability and lack of maintainers, Proxmox VE 9.0 drops native Gluster support. Recommended alternatives include:

• Ceph
• ZFS over iSCSI
• NFS or SMB (for archive/backup storage)

Users should migrate their volumes before upgrading.

Upgrade Path

Upgrade from VE 8.x

APT-based upgrades are supported, with a few precautions:

• Use the pve-test repo for now (stable will follow)
• Run pve8to9 to generate an upgrade checklist
• Review all breaking changes (permissions, storage, network)

Backups are strongly recommended before upgrade, especially for clusters or systems with custom SDN/firewall scripts.

Roadmap & Final Release

The stable release of Proxmox VE 9.0 is expected in Q3 2025, following extensive beta feedback and bug resolution. Further milestones include:

• Final Ceph Squid tuning
• SDN and HA refinements
• Documentation and translations update

Ideal Use Cases for Beta Testing

If you’re considering trying the beta, it’s most useful in:

• Lab environments testing SDN or Ceph
• Clusters with LVM or SAN storage needing snapshots
• Scenarios needing RAID-Z expansion in production ZFS setups
• Proxmox Backup Server compatibility validation

Proxmox VE 9.0 Beta is a bold step forward in infrastructure-grade virtualization. It modernizes key components like storage and networking while keeping the core Proxmox values: simplicity, transparency, and open-source power.

While it’s not yet production-ready, early adopters will appreciate the performance gains, better manageability, and future-proofing—especially in clustered or software-defined setups.