All case studies
Proxmox · Datacenter stretchedIT Contract Services · North-East Italy

New site, geo-distributed datacenter. Live migration: zero downtime for clients.

0 minutes of downtime during migration: no SLA violated
RTO < 5 minutes in case of complete site failure
Synchronous Ceph storage: RPO = 0 (no data lost on failure)

The challenge

A company providing IT contract services to end clients (hosting, management applications, corporate VPNs) needed to open a second operating site and transfer the existing infrastructure without interrupting services. End clients had contractual SLAs with guaranteed uptime: any downtime would generate penalties and reputational damage. Existing infrastructure was on single-site Proxmox — a great starting point, but without geographic protection.

The solution

We designed and implemented a Proxmox stretched cluster across two geographically separate sites, with synchronous Ceph storage replicated between both sites. Every write to the primary datacenter is replicated in real time to the secondary site — guaranteed <5ms latency between sites. Migration of existing VMs happened live, one at a time, with no downtime: each VM was live-migrated into the stretched cluster during end clients' normal working hours. The result is an active-active infrastructure: in case of complete site failure, VMs automatically restart on the other site in less than 5 minutes.

The results

0 minutes of downtime during migration: no SLA violated

RTO < 5 minutes in case of complete site failure

Synchronous Ceph storage: RPO = 0 (no data lost on failure)

Active-active infrastructure: both sites operating simultaneously

Infrastructure cost −40% vs commercial hypervisor alternative

The next case study could be yours.

Tell us your challenge. We always start with analysis — no commitment.

Call nowBook a consultation