What is vStack OS

vStack OS is the node operating system of the vStack HCP hyperconverged platform. It is a custom FreeBSD build: the base system, the kernel, and a set of controllers that turn an ordinary server into a cluster node. Every node runs virtual machines, storage, and networking at the same time; the cluster has no separate storage arrays or network controllers.

The cluster is managed by vStack Center. It runs in a management VM on the same cluster and has its own documentation: vStack Center. This documentation covers the layer below it: the nodes themselves, their installation, upgrade, and the utilities an administrator uses on a node.

What a node consists of

Layer	What it does	What the administrator sees
bhyve hypervisor	Runs virtual machines. The process of each VM is managed by the vstack-vm controller	The VMs of a pool run on the node that holds that pool
ZFS storage	Each pool is built as a raidz from one disk on every node. A node attaches its own disks locally and receives the other nodes' disks over iSCSI. If a VM runs on a node other than the owner of its pool, its disks are delivered over NVMe-oF/TCP	Pools are named z01, z02, and so on. A pool survives the loss of as many nodes as the redundancy level N+1…N+3 allows
Software-defined network	VM networks, isolated networks over VXLAN, external access points	Physically, these are two LACP aggregates, lagg10 and lagg20, with VLANs on top
Cluster Manager	The cluster framework: it monitors nodes, decides which node holds which pool, and moves pools when a node fails. On a node, this is the ccrd daemon	Each pool is represented by a zpool_<pool> service, for example zpool_z01. The cluster state is shown by the node CLI

In vStack Center this component has the same name, Cluster Manager, so the term does not change between the two documentation sets.

Who manages what

vStack Center does not touch nodes directly. When an administrator enables a node, disables it, or moves a pool, Center passes the command to Cluster Manager, which carries it out on the nodes. This has two consequences.

First, the cluster does not stop when Center stops. If the management VM is down, running VMs keep running, and Cluster Manager still moves pools when a node fails. What you lose is the web interface and the API, and along with them whatever brings up the VMs of a moved pool: starting machines after a pool failover is done by Center.

Second, the management VM itself lives on one of the cluster pools, usually z01, and has number 0 in that pool's VM list. You cannot move the pool with the management VM from the Center interface: Center would shut itself down in the middle of the operation. Such a pool is moved from a node, with the cluster utility. How to do this is described in the section on node upgrade.

Some roles fall outside this division. Node installation and the initial cluster assembly are done by the vStack OS installer; at that point Center does not exist yet. It deploys itself when the nodes boot for the first time. The node OS upgrade is also performed on the node itself, with the standard vStack.update.sh script.

Versions

The current version is 3.2.0 (25.09.2026), based on FreeBSD 14.2. Previous releases of the 3.x line: 3.1.0 (09.09.2026) and 3.0.11 (31.08.2026). Nodes of the 2.x line run on FreeBSD 13.2 and move to 3.x through a separate procedure, together with vStack technical support.

Documentation map

Section	What it contains
Hardware and network requirements	Processors, disk controllers and network adapters, storage disk layout, switching and VLANs
Node installation and cluster assembly	Installation from media, installer dialogs, first cluster boot
Upgrade to 3.2	Upgrading nodes one at a time with the standard script, checks, and rollback
Node configuration backup	A copy of the node configuration before an upgrade or reinstallation
Node CLI	The cluster utility and vStack.update.sh: commands, options, exit codes

Virtual machines, VDCs, networks, users, and roles are configured in vStack Center. This is described in its documentation.