Acceptable Use Policy

Effective July 21, 2026 · Part of the Terms of Service

01Principle

You get root, firmware-level control and privacy. In exchange, what you run must be lawful in the server's jurisdiction and must not damage the network, other customers, or third parties. We don't police content beyond that line — but the line is firm.

02Prohibited — zero tolerance

Child sexual abuse material, in any form. Immediate termination and report to authorities.
Outbound network attacks: DDoS, port-scan floods, spoofing, amplification, botnet command & control.
Malware distribution, ransomware operations, phishing sites and credential-harvesting infrastructure.
Unsolicited bulk email (spam) or SMS, and services sold primarily for it.
Terrorism-related content and anything under EU sanctions law.

03Restricted — allowed with care

Email sending — fine for transactional/opt-in volume with correct rDNS, SPF and DKIM. Sustained bulk sending needs prior approval so we can keep the IP space clean for everyone.
Security research — scanning or testing only against systems you own or are authorised in writing to test.
Public relays / exit nodes — VPN endpoints and relays are fine; public proxy/exit services that generate recurring third-party abuse must be able to answer their own abuse mail, or we'll route it to you and expect action.
Crypto mining — permitted; it's your power envelope. Sustained thermal abuse that trips platform limits may require a config change.
Adult content — legal adult content is permitted.

04Network fairness

Unmetered means unmetered — we don't bill traffic. We do expect traffic to be genuine service traffic: deliberately saturating the port 24/7 as a traffic-generation scheme, or reselling raw transit, degrades the shared edge and may lead to a port-speed conversation. Real workloads, backups, streaming, CDNs: push as hard as you like.

05Abuse handling — the process

Reports reach the abuse contact published in our WHOIS / RIPE registry records and are triaged by an engineer, not auto-forwarded to a ticket bot.
Ordinary complaints (a compromised service, a phishing page on a hacked CMS): we forward to your order email with a 24-hour window to respond or fix. Most incidents end here.
Active harm to third parties or the network: we null-route the affected IP first, then notify you. Service resumes when the issue is addressed.
Zero-tolerance categories (§02): immediate termination without refund, and lawful reporting where mandated.

06Fair warning

We terminate for what §02 lists, not for controversy, jurisdiction-shopping complaints, or DMCA notices sent to the wrong country. If a report has no legal basis where the server stands, we say so to the reporter — you'll be told, not sacrificed.