€100 welcome bonus a code for your next order, after your first renewal How the bonus works
Service level agreement, optional, +20%
Dedicated server SLAs usually stop at the network. Ours covers your server itself: its hardware, its network path and the time a failed part takes to swap, as a monthly availability figure in your contract, with compensation when we miss it. It costs 20% on top of the monthly, and we offer it because we are confident enough in the service to pay when it fails.
Not with a promise. With hardware nobody has to hope about and people who are already awake. We own and run the server, its spare parts and the network it is on, so this SLA is ours to keep, not another hosting company's passed on to you.
Two or more Juniper MX304 routers and two or more QFX10008 switches at every site, ten transit carriers, direct peering at 25+ exchanges. A failed link, line card, chassis or carrier is a re-convergence, not an outage.
EVPN-MPLS, own DWDM wavelengths, Nx100G per corridor
Scrubbing is on-net and always in the path: an XDP/eBPF filter in the NIC driver of the Sofia cluster, 4 Tbps and 1000M pps, a new rule enforced in under a second. An attack against you is measured under the same SLA. Keeping you up through it is our job.
On-net scrubbing, never a null route
The people who see the alert are the people who can change a route or swap a line card, 24 hours a day, every day. Nothing automated answers before a person does, and the spares are on our own shelf in the same building.
No first line, no AI in the escalation path
| Availability | Your server is reachable from our border on its uplink and can pass traffic to and from the internet. Measured per calendar month, per server. |
|---|---|
| Network | Our backbone carries packets between your port and our transit and peering edge. Loss on a third-party network beyond our edge is not ours to promise, but routing around it is. |
| Hardware | A failed component is replaced from our own spares by the engineer on site, day or night; the time the server is down for the swap counts as downtime. |
| Measurement | Our monitoring, sampled at least once a minute from inside our network, is the record. You may bring your own measurements to a claim; where the two disagree, we look at both. |
| Excluded | Planned maintenance announced to the technical contact ahead of time; downtime you cause (a kernel that does not boot, a firewall that locks you out, a reinstall); force majeure; suspension for non-payment or an acceptable-use breach. |
| Term | Added to any server at order time or later; removable at the end of any month. The option costs 20% of the monthly. |
Service credit as a share of that month's fee for the affected server, credited to your balance.
| Monthly availability | Downtime in a 30-day month | Service credit |
|---|---|---|
| 99.99% or above | up to 4.3 minutes | target met |
| 99.9% to 99.99% | 4.3 to 43 minutes | 10% |
| 99.5% to 99.9% | 43 minutes to 3.6 hours | 25% |
| 99.0% to 99.5% | 3.6 to 7.2 hours | 50% |
| below 99.0% | more than 7.2 hours | 100% |
Of the end of the month in question. Say which server and when you saw it down; our records do the rest. No form, no evidence pack required.
And answer with the measured availability for the month and the credit it maps to, or with the exclusion the time falls under, named.
On the next invoice for that server. Credits are capped at 100% of the month's fee per server and are not paid out in cash.
Bring your measurements. Where the two records differ we look at both, and the correspondence is with the engineer who handled the incident, not with a claims department.
Procurement that needs a signed availability figure, a regulator that wants to see one, or a customer contract whose own SLA you are backing with ours. The document is written for that reader.
A community game server, a side business, a service your friends depend on. You do not need a legal department to buy dependability. Tick the option and it is in the contract.
Certified and audited
Cloud Wall Ltd is certified for quality management and information security, compliant with PCI-DSS for card-handling environments, and operates within the European security and incident-reporting framework that applies to digital infrastructure providers under NIS2. GDPR applies to us both as a controller, for the account and billing records we keep ourselves, and as a processor for whatever runs on your server, and it is designed into what we retain.
Because the network and the engineers are the same either way. What the option buys is the contractual figure and the compensation, and we would rather price that honestly than fold it into every server for people who do not need the paperwork.
It is 4.3 minutes a month. The edge is doubled at every site, ten carriers sit behind it and the scrubbing is on-net. We publish the figure because we are confident in the reliability of the service, and we pay when we miss it.
Yes, from the next month. And you can remove it at the end of any month.
No. Keeping you reachable through an attack is what the on-net scrubbing is for, so an attack against your server is measured under the same SLA.
Tick the SLA when you order a server, or ask us for the SLA document before you sign.