What a service level commitment covers
A service level commitment applies to a defined system, over defined hours, measured in a defined way, with a defined remedy if it is missed. All four have to be agreed for the commitment to mean anything. Where any is absent, what exists is an aspiration rather than an SLA.
This framework applies to anything we operate under a support or managed arrangement, including offerings introduced in future. It does not apply to a system we built and handed over, where operation is yours, unless a separate operational agreement exists.
Availability
| Term | Definition |
|---|---|
| Availability | The percentage of covered minutes in a calendar month during which the covered service is reachable and performing its function |
| Covered hours | The period over which availability is measured — continuous, or defined business hours. Set per engagement ([TO BE CONFIRMED]) |
| Target | The committed availability percentage for the covered service. Set per engagement ([TO BE CONFIRMED]) |
| Measurement | Taken from monitoring we operate, at agreed check intervals from at least one point outside the environment |
| Exclusions | Scheduled maintenance, and the excluded events listed below, are removed from covered minutes |
Availability is a property of an architecture, not of a supplier's intent. A single server without redundancy cannot support a high availability target however carefully it is operated, and we will say so at design time rather than commit to a figure the design cannot deliver.
Response and resolution
Response time is the interval between a request arriving through the agreed channel and a person acknowledging it and beginning work. Resolution time is the interval to service being restored, which may involve a workaround where a permanent fix takes longer.
| Priority | Response target | Resolution approach |
|---|---|---|
| P1 — Critical | [TO BE CONFIRMED] | Continuous effort during covered hours until service is restored or a workaround is in place |
| P2 — High | [TO BE CONFIRMED] | Worked during covered hours until resolved or reduced in priority by agreement |
| P3 — Normal | [TO BE CONFIRMED] | Scheduled into normal work |
| P4 — Low | [TO BE CONFIRMED] | Addressed in the ordinary course, or in a scheduled maintenance window |
We commit to response times because they are within our control. We are cautious about committing to resolution times for faults whose cause is not yet known, and we would rather give a realistic estimate once diagnosed than a number at the outset that later has to be withdrawn. Priority definitions are set out in our support policy.
Scheduled maintenance
Maintenance is necessary and is planned rather than deferred. We give advance notice, agree windows that suit your operating hours rather than ours, and keep windows as short as the work allows. Notified maintenance within an agreed window is excluded from availability measurement. Emergency maintenance to address a security issue may be carried out with shorter notice, and we will explain why afterwards.
Exclusions
The following are excluded from availability measurement and from remedies. This is not an exhaustive list, and any additions are stated in the engagement documentation rather than introduced later.
- Failures in third-party services, carriers or connectivity outside our control, including upstream provider outages.
- Changes made by you or a third party to a covered system without our involvement.
- Failure of equipment, facilities or power that we do not operate under the engagement.
- Force majeure events.
- Suspension carried out in accordance with our terms, including for non-payment after notice or to contain a security risk.
- Any period during which we are unable to reach a system because access we depend on has not been provided.
Remedies
Where a committed target is missed for reasons within our control, the remedy is the one stated in your engagement documentation ([TO BE CONFIRMED]) — commonly a service credit calculated against the fees for the affected period. Service credits are the exclusive financial remedy for a missed service level, and are claimed by notifying us within the period the engagement documentation specifies. Nothing in this framework limits any right of termination for persistent material failure.
Reporting and review
Where an availability commitment exists, we report on it at an agreed interval, showing measured availability, incidents with cause and duration, and maintenance carried out. Where an incident has a cause worth acting on, the report says what will change rather than only what happened.
Service levels are reviewed when the architecture changes materially, because a target agreed against one design may be unachievable or unnecessarily conservative against another.
Contact
Questions about this document can be sent to info@kyconnects.com. We aim to respond to enquiries within 4–8 business hours.
Related documents
- Support Policy— priority definitions and how support operates
- Terms & Conditions
- Security Policy and Responsible Disclosure