Technical paper 01 · DR testing
Recovery you can prove
Every system, booted and tested every week. A test only passes when a real service answers, and every database responds.
Download PDF2 pages · 202 KBFor engineers who read the details.
Six short papers on how we approach backup, disaster recovery and ransomware, and what we measured on our own platform. Each paper stands alone, two pages, PDF.
Technical paper 02 · Disaster recovery
Same machine, different metal
Same virtual hardware, same MAC addresses, same VLANs, same clock. Why our DR copies boot without driver injection or re-IP.
Download PDF2 pages · 199 KB
Technical paper 03 · Backup integrity
A restore point is a promise
A restore point exists only after every byte is verified. Time-locked storage that nobody can release early, and no silent fallbacks.
Download PDF2 pages · 193 KB
Technical paper 04 · Failback
Coming home
Failback that sends back only what changed. The original keeps its identity and its backup chain.
Download PDF2 pages · 179 KB
Technical paper 05 · Ransomware
The last clean point
Every restore point checked for signs of ransomware. A clean restore that brings you back without losing the work since.
Download PDF2 pages · 185 KB
Technical paper 06 · Security model
Nothing listening
A fully managed service with real separation of duties: customers have no access to backup or restore, so there is no account to phish.
Download PDF2 pages · 178 KBTechnical deep-dive available under NDA.
Ask for the deep-dive For due diligence or a technical review, we share the full engineering version of each paper.
The papers are written in English.
Measured on our own platform.
A few numbers from the papers. Each one comes from a test on our own infrastructure.
2.17 Gbit/s
Failover traffic over our virtual DC-to-customer switch: an encrypted layer-2 path from our data centre onto the customer’s LAN. That was the full load in this test setup, not a limit: the speed also depends on the customer’s own connection and applications.
Paper 02300.1 → 7 GB
A forced full read, stored as an increment after deduplication, and restored byte-identical.
Paper 0334,000 files/s
An unchanged incremental of a share with 1.2 million files.
Paper 031.75 GB
Real file system data that a widely used “allocated areas only” shortcut left out in our testing. That copy booted into recovery mode.
Paper 03About ImmuBDR
The platform behind our backup and disaster recovery.
ImmuBDR is our own backup and disaster recovery platform, delivered as a fully managed service. It protects virtual machines, physical servers, databases and file shares, and recovers them on our own infrastructure, even when your own hypervisor or site is gone.
Proven weekly
Every system is booted and tested, with evidence you can show an auditor.
Immutable
Time-locked restore points that nobody can release early.
Own infrastructure
Storage, DR capacity, network and our own AI model, all run by us.
Managed by design
Our engineers run the platform. Every restore or failover is requested through a ticket, so your production and your recovery are held by different people.
Want to see a failover live?
Ask us for a demo. One of our engineers shows you a failover and the way back, on our own platform.