High Availability, Resilience, and Troubleshooting
Redundancy mechanisms
Know the HA options at each layer: VSF stacking (single logical switch, redundant members), VSX (two chassis with independent control planes and a sync link), MC-LAG, and gateway clustering. Understand that VSX requires an ISL (inter-switch link) and a keepalive path, and that split-brain prevention depends on the keepalive. At the wireless layer, understand gateway clustering and AP failover behavior.
Convergence
Spanning tree (MSTP), LACP, and routing protocols (OSPF, BGP) all affect convergence. Aruba recommends routed access to reduce STP dependence. Know that VSX supports active-active forwarding with fast failover.
Troubleshooting methodology
Use a layered approach: physical (cabling, PoE, optics), Layer 2 (VLAN, STP, MAC), Layer 3 (routing, ARP), and application. Tools include show commands, NAE, packet capture, and Central/AirWave diagnostics. For wireless, check RF (channel, power, interference), client (driver, roaming), and infrastructure (gateway reachability, role assignment).
Traps
- Confusing VSX (two control planes) with VSF (one control plane).
- Forgetting that VSX keepalive is separate from the ISL.
- Assuming STP is required in routed-access designs.
- Overlooking PoE budget exhaustion as a cause of AP/device failures.
Rule of thumb
For VSX split-brain: if the ISL fails but keepalive succeeds, one peer goes down to prevent dual-active. If both fail, both may stay active - design keepalive on an independent path.