Member OP
Updated incident report on the Strasbourg facility event now available. Our enterprise-grade redundant feeds maintained N+1 throughout the thermal event, though downstream capacity experienced temporary degradation. For affected parties, please contact sales regarding migration to alternate enterprise-grade zones. Full enterprise-grade restoration timeline available upon enterprise-grade consultation.
Member
Hey folks
Yeah we got hit bad on this one. Gonna be real, we had 40 boxes in that zone and our backup game was WEAK. Super stoked we moved half to OVHcloud last month or we'd be toast
Live and learn right? Now we got 3hr RPO max everywhere
Member
The Linux crowd often overlooks that Windows Server Backup with Azure integration handles offsite replication quite elegantly. I migrated our file services to KnownHost's Amsterdam zone six months prior, so we avoided the disruption entirely. Storage Replica has its merits for those considering hybrid approaches.
I run three Proxmox nodes across family members' fiber connections, sync nightly with Syncthing, and reverse proxy everything through Cloudflare. Docker Compose makes it trivial to spin up replacements. The fire was a wake-up call to own your stack.
"Our server is on fire please refund" — that was the ticket, and I had to explain that no, that was actually the facility, and no, our TOS covers Acts of God but not Acts of Datacenter. Got another one demanding we honor SLA during "temporary atmospheric conversion." Dry as the paperwork I'm filing.
Member
The Disaster recovery plan is not a Document; it is a Process. Please to implement the 3-2-1-Strategy with Offsite rotation and Tested restoration. I have moved all Prefixes to Hetzner with automated Prefix announcement switching. Please to verify your Backup integrity monthly.
Regards,
Klaus Mueller
2025-04-18T14:32:00Z
Member
Сервер in Strasbourg very slow now. Network not stable. I move to RackNerd, price good, network stable. Before I have no backup, now rsync to friend server in Poland. Learn hard way. Fire bad for business.
40 boxes with weak backups is rough
Member
Rough don't cover it. Lost a client who ran a booking engine for Quebec ski lodges, peak season too. Insurance guy asked if we had "geographic dispersion documentation." I made that up on the spot and he bought it.
Now I'm at OVHcloud Beauharnois + Hetzner Falkenstein. Ping sucks between them but whatever, one's on fire and the other ain't.
Member
Please to keep this Documentation with your Offsite rotation logs. The Insurer will deny Claims without Timestamped evidence.
I have updated my Runbook to include "facility thermal event" as explicit Scenario. Before was only "flood" and "cyber incident." Fire is now separate Category.
