A selective VLAN outage after a vPC change
A pair of NX-OS switches provides a dual-attached server access layer through vPC. Management access works to both peers. After an uplink maintenance task, one application VLAN stops passing traffic while another VLAN on the same bundle still works. The peer-link is operational, but the engineer has not yet inspected per-VLAN consistency or the allowed lists on both peers and the downstream device. A colleague proposes restarting both switches because a small management ping is successful and therefore the problem must be an application fault. The team must diagnose the selective forwarding failure without converting the investigation into an unrelated fabric-wide outage or assuming that carrier and management reachability validate every tagged service.
1. What read-only evidence should be gathered first?
Read the worked answer
Compare the intended and operational VLAN lists, port-channel member state, vPC state, and relevant consistency parameters across both peers and the downstream attachment. Trace the failing application VLAN and return path separately from the working management path.
2. What should guide the repair and verification scope?
Read the worked answer
Correct the identified VLAN or compatibility discrepancy using the approved targeted change. Confirm affected and previously working services afterward, and preserve rollback information. A global restart is not justified merely by management reachability or a selective VLAN failure.