Service Level & Availability
Our commitment on uptime, availability and planned maintenance.
This is a convenience translation. Only the German version is legally binding: German version.
10-11 am
1. Scope
This Service Level describes the targeted availability of the publicly reachable services of lionbackup, in particular the backup storage endpoints (data upload and download through the client) and the customer portal. It applies to production operation. The binding, contractually guaranteed values follow from the respective contract or the data processing agreement (DPA); in case of doubt they take precedence over the information on this page.
2. Availability (uptime)
We aim for a monthly availability of the storage endpoints of 99.9 %. Availability is measured as the share of a calendar month in which the respective service is reachable from outside and answers requests successfully, minus the excluded times listed below.
Calculation
Availability (%) = (total time of the month − unreachable time − planned maintenance − excluded times) ÷ (total time of the month − planned maintenance − excluded times) × 100.
| Monthly availability | Maximum unplanned downtime (guide value per month) |
|---|---|
| 99.9 % | about 43 minutes |
| 99.5 % | about 3 hours 39 minutes |
| 99.0 % | about 7 hours 18 minutes |
3. Planned maintenance window
A fixed, recurring maintenance window is reserved for maintenance, updates, security patches and deployments:
No planned maintenance takes place on Fridays or on public holidays in Baden-Württemberg. During this window individual or all services may be temporarily unreachable. They do not have to be: the window is the option to carry out maintenance without separate advance notice.
- No separate announcement: we carry out work within this window without individual announcement. It counts as communicated in advance.
- Not counted against availability: downtime within the maintenance window counts as planned maintenance and not as unplanned downtime in the availability calculation.
- Shorter than the window: as a rule the hour is not nearly used up; many maintenance tasks run without a noticeable interruption.
- Data security unaffected: data already backed up is not at risk during maintenance. Only access and the creation of new backups may pause.
Unscheduled (emergency) maintenance
In case of acute security vulnerabilities or serious failures we reserve the right to carry out maintenance outside the window as well. We keep such cases as short as possible and inform affected customers through the usual channel where reasonable.
Announced maintenance by our upstream providers
Our upstream providers (for example data centre and infrastructure providers) also maintain their systems. If an upstream provider announces work that affects our services, we carry out the measures this requires, such as shutting down and restarting affected systems in a controlled way, within the announced period. We announce such maintenance with reasonable notice on the page Wartungsarbeiten (maintenance, in German); like the fixed maintenance window, it counts as planned maintenance.
4. What does not count as downtime?
The following in particular do not count as unplanned downtime:
- work within the planned maintenance window (Mon to Thu, 10:00 to 11:00, only on working days in Baden-Württemberg);
- announced unscheduled or emergency maintenance;
- maintenance by our upstream providers (for example data centre and infrastructure providers) announced with reasonable notice, including the controlled shutdowns and restarts this requires on our side;
- failures due to force majeure (for example large-scale power or network outages, natural events);
- failures outside our sphere of influence (for example the customer's internet connection, DNS or routing problems of third parties, failures at upstream providers unless we are responsible for them);
- causes on the customer side (for example faulty configuration, expired credentials, exceeding agreed quotas).
5. Monitoring & alerting
All public endpoints are checked automatically around the clock for reachability, response time and valid TLS certificates. Threshold violations (for example unreachability, high latency, scarce storage) immediately trigger an internal alert so that our team can respond.
6. Support & response
You can report failures at any time through our contact form or our support. We accept reports and prioritise them by severity; the specific response and recovery times depend on the agreed support scope.