Skip to main content

ARP Roaming Issue Troubleshooting

Overview​

When wireless clients roam between access points, they may experience a temporary loss of connectivity to other devices on the LAN (typically lasting 2-5 minutes), while internet access continues to work normally. This is a common issue caused by switch MAC address table aging and is not related to the wireless system itself.

Why This Happens​

When a wireless client roams from AP-A to AP-B, the client's MAC address suddenly "moves" from the switch port feeding AP-A to the port feeding AP-B. However, most switches maintain a dynamic MAC address entry for the original port for a set period (typically 120-300 seconds).

During this aging period:

  • Internet traffic works: Outbound traffic immediately teaches the switch the new location
  • LAN traffic fails: Other devices attempting to reach the roaming client send frames to the old port location until the MAC entry ages out

The 2-minute connectivity blackout corresponds exactly to the switch's MAC aging timer.

Quick Diagnostic Checks​

CheckWhat to Look For
SSID ConfigurationVerify the SSID is not marked as "Guest" or has ACLs that block LAN access by design
Fast Roaming SettingsCheck if 802.11r/k/v is enabled (helps but doesn't fix the underlying switch issue)
Switch MAC Aging TimerUse show mac address-table aging-time or check Web UI. If you see 120-300s, this matches the outage duration
MAC Flapping LogsReview switch logs for "MAC flapping" entries when clients roam - this confirms the root cause

Fix and Mitigation Options​

1. Enable MAC Move Updates (Preferred)​

Configure the switch to accept rapid MAC address moves:

Cisco-style switches:

mac address-table move update receive

TP-Link switches:

  • Navigate to Switch > MAC Address Table > Enable Move Update
  • This allows the switch to immediately update MAC entries when APs notify it of client moves

2. Reduce MAC Aging Timer​

TP-Link switches:

  • Go to L2 FEATURES → MAC Address Table → Aging Config
  • Set aging time to 20-30 seconds instead of 120-300 seconds
  • Trade-off: Slightly higher CPU usage in very large networks, but acceptable for typical deployments

3. Network Topology Optimization​

  • Single Switch Deployment: Connect all APs to the same downstream switch
  • Avoid Cross-Switch Roaming: When AP-A and AP-B connect to different aggregation switches, MAC moves cross switch boundaries and take longer to propagate

4. Force ARP Cache Refresh​

  • Some WLAN systems send Gratuitous ARP (GARP) packets when clients re-associate
  • Client-side scripts can trigger DHCP lease renewal or send GARP to force LAN devices to refresh their ARP cache
  • This also prompts the switch to relearn the MAC address location

5. WLAN Profile Verification​

  • Disable Guest Mode: If LAN access is required, ensure the SSID is not configured for guest isolation
  • Consistent Configuration: Verify SSID, security settings, and VLAN configuration are identical across all APs
  • Mixed Profiles: Inconsistent settings between APs will break seamless roaming

Ubiquiti UniFi Equipment​

Does this issue occur with Ubiquiti equipment?

Yes, this issue can occur with UniFi deployments, but it's less common for the following reasons:

Why UniFi is Less Affected​

  1. Integrated Architecture: UniFi controllers communicate directly with UniFi switches, enabling better coordination of MAC address updates during roaming events

  2. RSTP and Fast Convergence: UniFi switches typically run optimized spanning tree protocols that handle topology changes more efficiently

  3. Controller-Based Management: The UniFi controller can coordinate roaming events between APs and switches in the same ecosystem

  4. Default Configurations: UniFi equipment ships with more roaming-friendly default settings compared to many traditional enterprise switches

When Issues Still Occur​

  • Mixed Vendor Environments: Using UniFi APs with non-UniFi switches
  • Legacy Switch Integration: Connecting to older switches without fast MAC learning capabilities
  • Large Layer 2 Domains: Very large networks may still experience occasional delays during roaming events

Recommendations for UniFi Deployments​

  • Use UniFi switches when possible to maintain ecosystem integration
  • Enable "Fast Roaming" in UniFi controller settings
  • Verify that switch ports connected to APs are configured for trunk/tagged VLANs appropriately
  • Monitor the UniFi controller logs for roaming events and any associated errors

Bottom Line​

The ~2-minute LAN connectivity loss after roaming is almost always caused by the switch holding onto the old port location for the client's MAC address. Enable MAC-move updates or reduce the aging timer, and the problem typically disappears immediately.

For mixed-vendor environments, focus on switch-side solutions. For UniFi-only deployments, ensure all equipment is running current firmware and that fast roaming features are properly configured.