Service Level Agreement
Last updated: June 10, 2026
99.5%
monthly uptime,
per server
per server
Most budget hosts promise 99.9% and pay nothing when they miss. We promise what a single well-run node can actually deliver and pay credits automatically when we don't. 99.5% allows ~3.6 hours of downtime a month - measured by our external monitoring, the same feed that powers the public status page. You see what we see.
Credits
| Monthly uptime | Credit (% of that server's month) |
|---|---|
| ≥ 99.5% | 0% - promise kept |
| 99.0 - 99.5% | 10% |
| 95.0 - 99.0% | 30% |
| < 95.0% | 100% - the month is on us |
Credits land on your balance automatically within 3 days of month end - no ticket, no begging. They pay for future service (not withdrawable).
Counts as downtime
- ✓ Your VM unreachable from the public internet because of our infrastructure: host crash, network, power, our DDoS-mitigation gone wrong
- ✓ Emergency maintenance without notice
Doesn't count
- ✕ Anything inside your VM: your OS, your firewall rules, your full disk (unmanaged plans are yours to run)
- ✕ Scheduled maintenance announced 48h+ ahead (we aim for under 30 minutes, monthly at most)
- ✕ Suspension for non-payment or AUP violations
- ✕ Force majeure: upstream fiber cuts, datacenter-wide power events beyond our redundancy
Honesty clause: every incident gets a public post-mortem on the status page - what broke, when, why,
what changed. If our monitoring disagrees with yours, send your logs. When in doubt, we side with the customer.