A vSAN means distributed, replicated storage built from the servers' own disks — no external SAN. VMware vSAN is commercial; here are the free options that do the same job and when to choose each.
What a vSAN solves
- No single point of failure — data is replicated across nodes; lose a node, the VMs keep running
- Hyperconverged — compute + storage on the same servers, no dedicated array
- Scaling — add nodes/disks and capacity grows
Free options
- Ceph (Proxmox) — most mature for 3+ nodes; 3x replica, object/block/file; integrated in Proxmox
- StarWind VSAN Free — excellent for 2 nodes (Hyper-V/ESXi/bare-metal); synchronous replication, a real free edition
- GlusterFS — distributed file storage, simple, good for large replicated volumes
- DRBD + Linstor — kernel-level block replication; low-latency for 2 nodes
- ZFS + replication — not a true vSAN, but snapshots + send/receive for DR on 2 nodes
How to choose
- 2 nodes — StarWind VSAN Free or DRBD/Linstor — simple synchronous replication
- 3+ nodes — Ceph — real HA, quorum, 3x replica (see the Proxmox+Ceph guide)
- Zero budget + simple — ZFS replication for DR, GlusterFS for files
- Network — at least a dedicated 10GbE for replication, whatever the solution
Conclusion
The right vSAN depends on node count and your RTO/RPO. For 3+ nodes we recommend Ceph on Proxmox; for 2 nodes, StarWind VSAN Free or DRBD. We size and deploy it for you.
Example: GlusterFS replicated volume (2 nodes)
The simplest SDS to start with — 2-node replication:
# GlusterFS — volum replicat pe 2 noduri (SDS simplu, tip vSAN)
sudo apt -y install glusterfs-server && sudo systemctl enable --now glusterd
sudo gluster peer probe node2
sudo gluster volume create vol1 replica 2 \
node1:/data/brick1 node2:/data/brick1
sudo gluster volume start vol1
sudo mount -t glusterfs node1:/vol1 /mnt/vol1
sudo gluster volume info vol1