Operations, RAMS and SLAs

After acceptance, evidence continues in operation

Talbahn's group call has been accepted. Months later, an authorisation expires and prevents its use. Servers are reachable and radio coverage exists, yet Lea and Niko cannot use the intended service.

A Service Level Agreement (SLA) is an agreed description of services and measurable performance conditions between parties. It must identify the service, measurement boundaries, assessment period and responsibilities. “Network availability 99.9 per cent” is insufficient if what counts as available is unclear.

An agreement for the actual purpose

A contractual exclusion does not make an operational impairment disappear. Talbahn should therefore display contractually excluded maintenance periods separately from actual unusable periods.

RAMS connects the operational questions

From 2.21 you know RAMS as a life-cycle framework. Its four perspectives now become concrete:

• Reliability: How often does the required function fail under the specified conditions?
• Availability: Is it usable when needed?
• Maintainability: How can it be checked, repaired and restored?
• Safety: Which risks arise, and how are they controlled in the responsible safety process?

EN 50126 provides the process framework, not an arbitrary universal FRMCS availability figure. High availability alone also does not prove adequate safety. [S15]

Common causes limit redundancy

Redundancy means providing additional means for the same required function. Two modems can help against failure of one modem. If they share power, antenna, configuration or service platform, those common dependencies can affect both paths.

Talbahn therefore maps the dependencies essential to its service. It checks spare parts, recovery times and support availability. Software changes and renewed credentials must also be planned.

A cryptographic key is a technical value used in a protection mechanism. Suitable keys can, for example, enable verification of identity evidence. In technical security, a certificate can confirm the binding of an identity to such a key. Its expiry can block an intended registration or connection. Managing these lifetimes therefore belongs to operations; this certificate differs from a course attendance certificate.

A lead owner prevents responsibility gaps

Here, a Service Owner is the named party responsible for the service across the technical areas involved. It need not operate every component itself, but coordinates requirements, changes and incidents.

Keep shared service dependencies visible

The operating overview lists SIP Core, MC services, identity, configuration and key management, and required gateways. For each function record which new and ongoing activities its failure affects and how recovery is checked. A second radio connection does not automatically improve a shared dependent service function.

Check your understanding

A supplier meets its server-availability target while expired certificates prevent the group call. Is the operational requirement fulfilled? What do you add?

Model answer

No. The server metric does not fully represent service usability. I add end-to-end measurement, identity and certificate management, renewal responsibility and a tested recovery procedure. Contractual assessment of the server component and the operational service effect are reported separately.

What to remember

Sustained service quality needs measurable boundaries, maintained dependencies and an accountable owner.

Your Talbahn case file

Add the service boundary, incident leadership, certificate management and common failure causes to your operational agreement.