Symptom

Clients are stuck on ?mobility? state in anchor. I see this message as well on console. May 20 17:15:55.487: %MM_LOG-4-RETRIES_FAILED: Chassis 2 R0/0: mobilityd: MAC: c0:56:27:bc:c3:b7: All retries of export_anchor_req (XID: 894) to ipv4: 9.8.45.60 have been exhausted May 20 17:15:55.487: %MM_LOG-6-MM_EXPORT_ANCHOR_FULL: Chassis 2 R0/0: mobilityd: Export Anchor not responding, back off condition set (Client MAC: c0:56:27:bc:c3:b7, AnchorIP: IP: 9.8.45.60) I have tried using both default and named profile policy ? I get the same result. Issue is seen with both IPv4/IPv6 tunnels. Tried associating new clients(to negate client issue) - still the same result. eWLC#clients Number of Clients: 2 MAC Address AP Name Type ID State Protocol Method Role ------------------------------------------------------------------------------------------------------------------------- 8c85.90f1.87ac AP-3800 WLAN 9 Mobility 11ac MAB Unknown c056.27bc.c3b7 AP-3800 WLAN 9 Mobility 11n(5) MAB Unknown Anchor config: WLC#sh run | sec cwa wireless profile policy cwa aaa-override mobility anchor nac no shutdown Foreign config: eWLC#sh run | sec cwa wireless profile policy cwa aaa-override mobility anchor 9.8.45.60 priority 3 nac no shutdown Other outputs and logs attached to the CDET.

Conditions

CAPWAP Mobility message getting fragmented on the length of a mobility payload.

Workaround

Change the length of the REDIRECT ACL or URL. This will avoid fragmentation on the length field. However, clients who tried to join anchor will be stuck. A reboot on guest anchor controller is needed to clean up the stale clients.

Further Problem Description