PROXMOX & CEPH•2026-10-10•1 MIN READ
High Availability Proxmox VE 8 Cluster with Ceph Storage
Step-by-step configuration of a 3-node enterprise hypervisor cluster with shared Ceph block storage and redundant Corosync.
High Availability Proxmox VE 8 Cluster with Ceph Storage
Building a quorum-backed 3-node virtualization cluster allows virtual machine live migration without 1 millisecond of packet loss or downtime.
Quorum Safety Guideline
Never build a two-node cluster without a low-cost QDevice Raspberry Pi witness to avoid split-brain failure scenarios.
1. Initialize Corosync Cluster Link
Run the initialization sequence on node 1 with dedicated 10GbE corosync ring links:
Corosync Cluster Initialization
Proxmox 8.2
$
pvecm create technhit-cluster --link0 10.10.10.12. Multi-Node Corosync & Ceph Network Architecture
Below is our redundant network map:
3-Node Corosync & Ceph Storage Topology
SVG // HIGH-AVAILABILITY TOPOLOGYPVE-Node-01 (10.10.10.1)Corosync Link0 + Ceph OSD
PVE-Node-02 (10.10.10.2)Corosync Link0 + Ceph OSD
PVE-Node-03 (10.10.10.3)Corosync Link0 + Ceph OSD
3. Join Subsequent Hypervisor Nodes
On nodes 2 and 3, join the cluster:
Join Cluster
Proxmox 8.2
$
pvecm add 10.10.10.1 --link0 10.10.10.2Verify that quorum is active across all voting nodes:
Verify Quorum Status
Proxmox 8.2
$
pvecm status