Aruba Mobility Expert (Written) study materials
Aruba Mobility Expert (Written) study resources

OSPF Adjacency Flaps After Controller Uplink Change

Aruba Mobility Expert (Written) · Put your knowledge to work in context.

All case studies

OSPF Adjacency Flaps After Controller Uplink Change

A Mobility Controller's uplink was moved from an access switch to a new distribution switch to increase bandwidth. The controller runs OSPF to advertise its management and client VLANs into the routed core. After the move, OSPF adjacencies flap every few minutes, client VLAN routes intermittently disappear, and some clients lose gateway connectivity briefly. The new distribution switch is a different vendor and uses a non-default OSPF hello/dead interval. The controller's uplink is a single 10 Gb port with no LAG.

1. Explain the likely cause of the OSPF flapping and how to resolve it.

Read the worked answer

OSPF requires matching hello and dead intervals on both ends of an adjacency. The new distribution switch uses non-default timers, so the controller and switch never form a stable adjacency - hellos are missed, the dead timer expires, and the adjacency repeatedly tears down and reforms. Each teardown withdraws the controller's advertised VLAN routes, causing the intermittent gateway loss.

Resolution:

  • Align OSPF timers (hello/dead) on the controller interface to match the distribution switch, or change the switch to standard timers.
  • Verify area, network type, and authentication match.
  • Confirm the interface is not flapping at L1/L2 (single 10 Gb port, no LAG - check for errors/CRC).
  • After alignment, confirm the adjacency stays up and routes remain stable.

2. Beyond timers, what redundancy improvement would you recommend for the single 10 Gb uplink?

Read the worked answer

The single 10 Gb port is a single point of failure and offers no L2/L3 redundancy. Recommend:

  • LAG (link aggregation) across two ports to the distribution switch (or to a pair of switches with MC-LAG/vPC) to provide link redundancy and bandwidth.
  • Ensure OSPF runs over the LAG and that BFD is considered for fast failure detection (with timers aligned).
  • Validate failover by disabling one member link and confirming the adjacency and client routes remain stable.

This removes the single-link risk that compounds the timer issue.