Survivorship bias. The thousands of other $3 customers who got the nod from support to migrate or got booted for TOS violations aren't posting. You're the one left standing, not evidence the model works.
My solar rig has a similar problem. "It's been up 400 days!" Yes, because I sized everything triple and I got lucky on a mild winter. Doesn't make it engineering.
3-2-1 rule, people. One of those copies is on Contabo, fine. Where's the offline copy? Air-gapped? Tested?
I don't care about your uptime. I care about your last restore test. When did you actually unpack that rsnapshot and verify the database imports clean?
Never? Then you have zero backups. You have backup theatre.
Member OP
October 2024. Full restore to a BuyVM slice in Las Vegas, verified in browser, let it run a week, killed it. Database imports clean, wp-config needed fresh salts but that's expected.
Not monthly like I should. Not never.
I'm on the same RackNerd Bronze plan, different DC — Chicago instead of whatever jane's on. 847 days here. Two unplanned outages, both under 15 minutes, both at 3am US time which is prime admin window so probably maintenance they didn't announce.
Ticket response last month: 11 hours. Not 4-8.
The DDoS protection they advertise is just Voxility remote-triggered. I know because I got null-routed during a small attack and the ticket explanation was "Voxility activated." Took 6 hours to lift.
Works for my price. Wouldn't put a client on it.
Member
This is precisely my concern. Under GDPR Article 32, "ability to restore availability" is a security requirement. Six hours of unavailability with no communication — if you were processing EU personal data, that's a reportable incident.
RackNerd's Amsterdam presence does not absolve them. Their US infrastructure handling EU data subjects triggers the usual transfer mechanisms, and I see no evidence of updated SCCs post-Schrems II.
I requested their DPA in February 2024. No response. I would not use https://www.racknerd.com for anything touching EU individuals.
I had RackNerd for 14 months. Three "network maintenance" windows with zero notice, one of which killed a RAID rebuild I had running. Lost the array.
Their status page showed green the whole time. I opened a ticket, got "we experienced unplanned network events in LAX" three days later.
Migrated to Hetzner. Paying 2x, sleeping 10x better.
Member
Most of my Madrid clients are on OVH Gravelines or Scaleway Paris for exactly this. The DPA paperwork is boring but it's there. I can show an auditor.
That said, I keep a RackNerd sandbox for testing WordPress updates. €3 to not risk a client site is fine. It's not production, it's a disposable hammer.
This thread made me check my own "zombie" host — Ionos 1&1 legacy plan from 2017. Still billing me £1.99 somehow, though their current cheapest is £4. I never migrated the domain because the auth code email goes to a dead address.
So I'm paying £24/year for a holding page and a domain I can't transfer. Who's the zombie now?
Same here but with Namecheap shared hosting from 2015. Auto-renews on a card I keep meaning to cancel. Hosts a single JPG that gets 400 hits a year from some ancient forum signature.
The real crime is I once needed that JPG and couldn't find the login. Support reset it in 20 minutes though. More than RackNerd would manage probably.
London perspective: RackNerd doesn't have a UK datacenter, so any EU data I put there is leaving the adequacy decision zone to the US. Even with SCCs, the US CLOUD Act exposure is real.
I use Krystal for UK clients, Mythic Beasts for the paranoid ones. Both answer tickets, both have working status pages, both cost 3-4x RackNerd.
You get what you pay for, but sometimes what you pay for is "not having to explain to the ICO why my processor vanished."
