I currently have HA running on two controllers (locals) on dual mode (preemption enabled) and LMS enabled with the default 600 second timer. We have a couple of sites with the same setup and in most cases the AP's will stay on the primary but have a few sites where AP's fail-over to the backup local and never revert back to the primary.
I checked the HA heartbeat counters and don't show anything missed between controllers. I also tried on manually moving the AP's (ap-move) to the primary LMS and they move back to the backup. The network is stable between to the two controllers that should make the AP's revert back to the primary. Is there other atturbites that are taken into account before the AP's reverts back to the primary controller?
AP system profile "AP_SYS_PROF"
Parameter Value
LMS IP 10.0.0.1
Backup LMS IP 10.0.0.2
LMS Preemption Enabled
LMS Hold-down Period 600 sec
HA group information
Parameter Value
Preemption Enabled
Over-subscription Disabled
State Synchronization Enabled
Pre-shared Key *****
Inter Controller heartbeat Enabled
Heartbeat Threshold 5
Heartbeat Interval 100
HA group-member IP address 10.0.0.1 dual
HA group-member IP address 10.0.0.2 dual
HA group-member IPv6 address N/A
HA Failover Information
Date Time Reason (Latest 10)
2017-07-25 12:29:54 Pre-emptive failover back to LMS 10.0.0.2
2017-07-25 12:19:54 AP move request received from controller: fail-over to 10.0.0.1
2017-07-14 19:51:08 Missed heartbeats: Last Sequence Generated=61372 Sent=20757 Rcvd=61363; eth Sent=2965713 Drop=0; gre Sent=2748348 Drop=0 First=1; tun Sent=0 Drop=0: fail-over to 10.0.0.2