I have a pair of WLC200R in a cluster. One is primary seed and the other is secondary. The primary has all the AP, radio profiles, and service profiles defined. The secondary has none of these until it joins the cluster - at which time it gets a copy of the radio and service profiles. However it doesnt list any ap definitions.
In the event that the primary member goes down, how does the secondary take over? Does it start using the primary's ip addresses? How does it service the APs?
Sounds like it could be a licensing issue. Have you ensured that your secondary WLC is licensed for the additional AP's and High Availability? The show license command will give you the number of licensed AP's for that platform, licensed number of redundant AP's and additional features such as High Availablity.
Does that mean that the secondary seed doesn't show the AP definitions on issueing
sho conf cluster area ap
Apart from that, in a properly working cluster the APs maintain connections to two controllers ("sho cluster ap" will tell you which ones are PAM / SAM). If the primary fails the AP will seamlessly switch over to the secondary, normally within 150ms. This does NOT involve a reboot of the AP.
In case the AP gets rebooted for other reasons while a controller is down it's a matter of how your APs locate controllers. If they're not on the same subnet and the corresponding DHCP options are not defined then it's up to DNS resolution. If the entry exclusively points to the failed controller the AP obviously won't find the working one. In this case you'd have to point the DNS record to the surviving MX.
But as mentioned: this only applies to APs which get rebooted after failover, the failover it self will work without intervention.
I experienced this same error message and it appeared on a member not a seed. It turns out that I had a miss-spelled VLAN on the member that wouldn't sync. After correcting it the member synced to the rest of the cluster. Hope this helps.