Separate neighbor state from route behavior
Identify the peer, local and remote autonomous systems, address family, routing context and affected prefixes. State whether the session never establishes, repeatedly changes state or remains established while the service fails. Each symptom calls for different evidence.
Record the intended route behavior before examining deviations. Cisco’s BGP troubleshooting material separates neighbor establishment, missing routes and advertisement problems; use that distinction to keep the investigation focused.
Collect evidence that answers the routing question
Use platform-appropriate read-only checks and preserve the observation time. Where remote-peer evidence is missing, make the uncertainty explicit.
| Symptom | Evidence to review | Avoid assuming |
|---|---|---|
| Neighbor does not establish | Transport reachability, peer settings and relevant logs | That a protocol configuration error is the only cause. |
| Neighbor flaps | Session events aligned with link and resource observations | That one snapshot explains an intermittent failure. |
| Prefix is missing | Received, accepted and installed route evidence where available | That an established session carries every intended prefix. |
| Unexpected path | Relevant policy and competing path evidence | That one attribute explains the full forwarding outcome. |
Illustrative example: the session is up but the prefix is absent
A service becomes unreachable after a reviewed network change. The BGP session remains established, but the engineer needs to determine whether the required prefix is received, accepted and installed in the relevant context.
The configured workflow gathers those observations and relevant change context. It reports the point where evidence diverges from the intended design. A route observation should be checked against the service path before it becomes a confirmed root cause.
Use NetPrompt with routing-specific context
Supply the relevant Network Context and diagnostic guidance, then validate the commands for each platform and version. Keep the first workflow scoped to one peer or prefix rather than requesting an unrestricted network-wide repair.
Template-only command selection can help bound the checks where configured. Missing command coverage should produce an explicit gap. Avoid presenting a generated command as an independently validated platform procedure.
Review proposed changes and recovery
A session clear, routing-policy edit or route-preference change may affect more than the reported service. Assess scope, require the applicable approval and define the post-change checks before execution.
Evaluate the workflow with a known incident, a healthy peer and incomplete remote evidence. The expected result is a reproducible explanation with relevant observations, unanswered questions and a service verification plan—not a guaranteed automatic fix.
Common questions
- Can BGP be established while a service is unreachable?
- Yes. Session establishment and the presence of the required usable route are different questions. Verify routing policy, forwarding and the service outcome.
- Can this cover multiple router vendors?
- Use the relevant device profiles and command templates, and validate each software version and diagnostic task during evaluation.
- Will NetPrompt automatically clear BGP sessions?
- This use case starts with diagnostic evidence. A disruptive action requires separately reviewed scope and the applicable execution controls.
Continue exploring
Further technical reading
External references explain general practices; they are not endorsements of NetPrompt AI.