Please review the FglAm Guide for Configuring the Agent Manager to work in HA mode.
The concept is the following:
How fail-over works:
The initial fail-over of agents from a Primary to a Standby is automatic.
An alarm is generated indicating the Fail-Over is in progress.
When the 'standby' FglAm then hosts the agents, they remain with the 'standby' until the user takes action to stop the FglAm which in turn then initiates the fail-over back to the original FglAm.
The attached groovy script can be run using fglcmd or within the Script Console.
In the example below, a fail-over has occurred as designated by the mState parameter. The previous Primary FglAm is now in STANDBY mode.