Availability and SLA
The UP-NETWORK SLA page defines the levels, calculation and exclusions. Your service’s offer, quote or contract specifies the applicable commitment.
Three levels
Best Effort
VPS BASIC, VPS PRO, web services and UP-Transport
No contractual percentage or automatic compensation.
Guaranteed SLA
VDS, dedicated servers, UP-Connect and Housing
99.9% per calendar month, with UP-NETWORK monitoring required.
Contractual SLA
Architecture validated by quote
99.95% if expressly provided for, with contract-specific terms.
The Collaborative Suite also advertises 99.9% on its dedicated page; check the terms of its offer.
Calculation and compensation at 99.9%
Availability is calculated over the calendar month: (length of the month − eligible downtime) / length of the month × 100.
Over thirty days, 99.9% corresponds to 43 minutes and 12 seconds of eligible downtime. Compensation begins only if the result falls strictly below 99.9%.
Once the threshold is crossed, the month’s eligible downtime periods are aggregated. Each started hour of ordinary downtime or started two-hour period of urgent maintenance gives one day of service, according to the published method, capped at thirty days. Thus, fifty ordinary minutes give one day if the monthly threshold has been crossed.
Scope and exclusions
For an Unmanaged VDS, the scope covers the infrastructure up to VM availability. Managed adds the components explicitly administered by UP-NETWORK. Client-managed applications and changes remain outside this scope.
Planned maintenance is excluded with at least twenty-four hours’ notice; urgent maintenance enters the calculation according to the published terms. Interruptions caused by DDoS are excluded. The second SwissShield Permanent IPv4 address has no availability SLA.
The required monitoring and other exclusions are detailed on the official page. A 99.95% SLA does not automatically use the 99.9% compensation arrangements. An ordinary outage affecting a supplier chosen by UP-NETWORK remains attributable under this policy.
For an incident, open a ticket with the times and observed impact.
Updated on 3 October 2026. Reference: SLA policy.