An enterprise satellite service should be evaluated against the business work it must sustain. Turn terms such as carrier grade, priority and resilient into written performance definitions, recovery responsibilities and an acceptance test. The most useful comparison is between complete service offers for the same sites, applications and operating hours.
Define a measurable service boundary
Specify the endpoints: the customer-facing Ethernet port, a provider gateway, a private-network destination or the application itself. A provider can measure its network as available while a local router, power supply or authentication service prevents useful work. Put ownership of each component beside the network diagram.
List essential applications, concurrent sessions, normal and fallback operating modes, and the longest tolerable interruption. Give each application an observable success criterion: a completed transaction, an intelligible call or a synchronized record. Ask the supplier to demonstrate the intended VPN, addressing and routing arrangement.
Separate a committed information rate from a peak or best-effort rate. Ask when and where the commitment applies, what traffic is eligible and which limits remain during congestion. Prioritized traffic is a scheduling policy; the contract determines whether any throughput or availability is guaranteed.
Read the definition behind the percentage
Scroll the table horizontally on small screens. Keyboard users can focus the table region and use the arrow keys.
| Clause | Question to resolve | Evidence to retain |
|---|---|---|
| Availability | What counts as unavailable, and at which endpoint? | Measurement method, interval and eligible equipment. |
| Exclusions | How are maintenance, obstruction, local power and customer changes treated? | The complete exclusion list and responsibilities. |
| Performance | Are latency, loss, jitter or throughput committed? Under what load? | Directions, test endpoints, time window and thresholds. |
| Restoration | Who acknowledges, diagnoses and repairs an incident? | Support hours, escalation route, replacement logistics and targets. |
| Remedy | How are credits claimed and calculated? | Claim deadline, supporting logs, caps and eligibility. |
As an example of why the details matter, Starlink's published SLA explanation describes eligible Priority plans, terminal eligibility, an outage-counting rule and service credits. Read the current terms attached to the actual order. A credit provision defines financial compensation; the continuity plan must address the business consequences of an interruption.
Find shared failure points
Draw the primary and backup paths from the user's device to the application. Highlight common power, routers, firewalls, cable routes, gateways, support teams and software dependencies. Two links attached to one firewall share that firewall as a failure point.
Ask which hazards the design is meant to survive. A satellite path can diversify a local terrestrial access route, but still relies on working site equipment and the provider's ground infrastructure. If both paths depend on the same building power, provide a tested power arrangement appropriate to the required runtime.
Record reduced-service expectations. A backup may carry essential transactions and voice while bulk transfers pause. Configure and test that behavior deliberately rather than allowing every workload to contend for the smaller path.
Run an application-level acceptance test
- Establish the baseline. Record firmware, configuration, provider plan, routes, normal load and timestamps.
- Test the workload. Measure both directions and complete the essential applications during representative busy periods.
- Exercise recovery. In an approved window, withdraw the primary path by the agreed method. Measure detection, route change and usable application recovery separately.
- Inspect dependencies. Check DNS, VPN renegotiation, address changes, multifactor login, inbound services and ongoing sessions.
- Test the return. Restore the primary path and observe failback, session disruption and route stability.
- Repeat with realistic load. Verify that deferred jobs remain deferred and essential work stays usable.
Cisco's video QoS guidance discusses bandwidth, delay, jitter and loss as contributors to call quality. Use the actual application's requirements and measured outcomes; local traffic priority cannot create additional satellite capacity or eliminate propagation time.
Distinguish a fast route change from a fast recovery
Record whether existing sessions survive or must restart. A successful new browser session says little about an interrupted upload or a persistent application connection. Agree who fixes each failed test and which evidence is required before acceptance.
Keep the evidence useful after installation
Download the enterprise satellite acceptance checklist. Store the signed service definition, configuration references, test results and escalation contacts together. Review them after firmware, routing, provider, power or business-application changes.
For networks spanning schools or branches, continue with multi-site planning. For equipment-level observations, use the satellite modem diagnostic guide.
Sources and further reading
Source links reviewed September 28, 2026. Check current service terms, equipment documents and operational notices when applying the guide.