Proxmox 9.2 closes the last major gap with VMware vSphere
For those who still had doubts: Proxmox is now a full-fledged enterprise alternative.
In the past, there was one feature IT administrators cited as a reason to stick with VMware vSphere: the Distributed Resource Scheduler (DRS). This system automatically redistributes virtual machines across cluster nodes based on real-time load. Proxmox didn't have it. Administrators had to intervene manually or work with scripts and third-party tools that only partially accomplished what DRS had been doing for years. For small environments, this was acceptable. For medium and large clusters, where hundreds of VMs run and workloads constantly fluctuate, the lack of automatic rebalancing was a serious operational burden. That hurdle has now been removed.
With the release of Proxmox Virtual Environment 9.2, there is now a full-fledged alternative.
The Dynamic Load Balancer: open source DRS
The most important new functionality in Proxmox VE 9.2 is the new Dynamic Load Balancer, integrated into the Cluster Resource Scheduler (CRS). While previous versions only supported static scheduling – a VM was assigned to a node and remained there unless you intervened – the CRS now actively monitors CPU and memory load across all cluster nodes.
If a node becomes overloaded while another is relatively free, Proxmox automatically migrates the workloads, respecting the HA rules you've configured. No manual intervention, no late-night balancing rounds, no scripts doing half the job.
This is exactly what VMware administrators know as DRS. And it's now built into the Proxmox GUI. No add-on, no license, no external tool.
What else is in 9.2
The Dynamic Load Balancer is the star of this release, but 9.2 brings more:
Extended SDN stack
WireGuard and BGP have been added as new fabric protocols. In combination with detailed BGP/EVPN filtering via route maps and prefix lists, Proxmox now offers serious software-defined networking for complex environments.
Disarm and Arm for cluster maintenance
New workflows for scheduled maintenance. "Disarming" a node before working on it, and then safely "arming" it again without unwanted HA triggers. Solid and long-requested.
Custom CPU models via the GUI
Management of custom CPU profiles moves entirely to the web interface. Useful for migrations between hardware with different CPU generations.
Modern Platform Stack
Debian 13.5 "Trixie", Linux kernel 7.0, QEMU 11.0, ZFS 2.4, and Ceph Tentacle 20.2 as the new stable standard.
What this means for those still on VMware
If your organization is still running on VMware and re-evaluating licensing costs after the Broadcom acquisition, now is the time to seriously consider Proxmox.
The technical gap has been closed. The functionalities enterprise administrators need (HA, automatic load balancing, SDN, Ceph integration, central GUI) are all present. What remains is careful migration planning.
Epact can help you with this transition: from initial evaluation to production environment. Contact us if you want to know what a migration from VMware to Proxmox means for your infrastructure. Completely non-binding and objective.