sounds like the similar problem in our scenario:
8540 WLC HA SSO Cluster
- RP Link drops/disconnects
- standby WLC reboots due to power cycle or command line
- should-be-standby WLC comes back active after waiting 120 seconds for peer to respnd
- at this moment both WLCs are active and connected VLANs are disturbed which causes network errors even for clients in WLANs attached to normal operational WLCs on the same VLANs.
v8.5.131.6 resolves the described behavior and reboots the standby WLC to maintenance mode.
