Network & Infrastructure / BGP Policy & Internet Edge
Graceful restart and stale routes during a forwarding failure
A router's control plane restarts and its peers retain routes, but the forwarding plane has also failed. Traffic blackholes while routing appears intentionally stable.
Environment and inputs: BGP speakers and route reflectors; policy syntax, RPKI integration and address-family behavior depend on the implementation. Diagnostic/design example; not tested on production equipment.
Collect the evidence
Correlate restart signaling, stale-route timers, line-card health and packet forwarding on the restarting device. Record whether the platform can actually preserve forwarding during the tested restart type.
Explain the result
Graceful restart assumes specific forwarding continuity. Retaining stale paths is harmful when that assumption is false; protocol continuity and data-plane continuity must be evaluated separately.
Plan the fix
Align graceful-restart participation and timers with supported platform behavior and the failure detector design. Avoid maximizing retention just to reduce route churn during maintenance.
Check the result and rollback
Test control-plane-only and complete-device failures separately. Require application recovery within the agreed bound and revert retention changes if stale next hops outlive usable forwarding.